shockbot/modes.ts

// changes to mode definitions should be reflected in docs/modes.mdx
import { REVIEWER_AGENT_NAME } from "./agents/reviewer.ts";
import { type AgentId, formatMcpToolRef, pullfrogMcpName } from "./external.ts";

export interface Mode {
  name: string;
  description: string;
  // step-by-step guidance returned when the agent calls select_mode.
  // custom user-defined modes supply this; built-in modes define it here.
  prompt?: string | undefined;
}

// Default user-facing summary format embedded in Review mode review bodies.
// Deliberately scoped to Review (initial PR review). IncrementalReview keeps
// its own terser bullet-list "Reviewed changes" shape since re-review bodies
// are deltas, not introductions. Distinct from the agent-internal snapshot
// (action/utils/prSummary.ts) which has its own stable scaffold and is never
// shaped by user instructions — see selectMode.ts for the firewall.
export const PR_SUMMARY_FORMAT = `### Default format

The body has at most three parts in this exact order:

1. **Reviewed changes preamble** — one bolded inline lead-in describing what the PR does, then a bullet list of the substantive changes, then a collapsed \`<details>Review metadata</details>\` block.
2. **Cross-cutting issue sections** (zero or more) — one \`### \` heading per concern, with a human-readable problem write-up and a collapsed \`<details>Technical details</details>\` block underneath.
3. **\`### ℹ️ Nitpicks\`** at the very bottom (only if there are nits worth surfacing in the body) — a flat bullet list, no technical-details block.

Inline-vs-body discipline is the most important rule: any concern that anchors to a single line in the diff goes in an INLINE comment via the \`comments\` parameter, never the body. The body is reserved for **cross-cutting** concerns — design-level issues that span files, meta-patterns across endpoints, or symptoms whose root cause isn't on any one line. If everything you found can be inlined, the body has zero \`### \` issue sections — just the preamble + metadata.

## 1. Reviewed changes preamble

Open with a single bolded inline lead-in followed immediately by the bullet list (no \`### Key changes\` heading, no \`<b>TL;DR</b>\`):

\`\`\`
**Reviewed changes** — one sentence on what the PR does and why. Focus on intent, not mechanics.

- **Short human-readable title** — 1 sentence per substantive change. Write a short prose phrase; when you name a file, type, or function, put that name in backticks (e.g. **Add \\\`TodoTracker\\\` for live checklists**). A reviewer should understand the full PR from this list alone — this IS the dispassionate "what was reviewed and what changed" overview, so cover the substantive changes, not just the loudest ones.

<details><summary>Review metadata</summary>

- **Mode:** Review (initial)   *or*   IncrementalReview (delta against prior pullfrog review)
- **Files reviewed:** {file_count}
- **Commits reviewed:** {commit_count}
- **Base:** \\\`{base_ref}\\\` (\\\`{base_sha_short}\\\`)
- **Head:** \\\`{head_ref}\\\` (\\\`{head_sha_short}\\\`)
- **Reviewed commits:**
  - \\\`{sha_short}\\\` — {commit_subject}
  - ...
- **Prior pullfrog review:** none   *or*   \\\`{prior_sha_short}\\\` (linked to the prior review URL when available)
- **Submitted at:** {iso_timestamp}

If \\\`HEAD\\\` of \\\`{head_ref}\\\` has advanced past \\\`{head_sha_short}\\\`, this review may be partially or fully stale — re-diff against \\\`{head_sha_short}\\\` before treating any technical-details block as current.

</details>
\`\`\`

Pull every metadata field from the \`checkout_pr\` tool's response — file count, commit count, base/head ref + SHA, the commit list. For \`IncrementalReview\` runs, populate \`Prior pullfrog review\` with the prior review's commit_id (short SHA) and link to its URL via \`list_pull_request_reviews\`.

## 2. Cross-cutting issue sections (zero or more)

For each cross-cutting concern, one \`### \` section. Use this exact shape:

\`\`\`
### {emoji} {short, descriptive title — what's wrong, not what to do}

{Human-readable problem write-up. Describes the PROBLEM only — what's broken, what the symptom is, what the blast radius is. NO asks, NO suggested fixes, NO "the right thing to do is...". Asks and fixes live in the technical-details block below; the visible part is for the human to *understand* the problem, not to implement it.}

<details><summary>Technical details</summary>

\\\`\\\`\\\`\\\`markdown
# {title repeated}

## Affected sites
- {file path:line} — {what's wrong there}
- ...

## Required outcome
- {what the fix needs to achieve, not how to achieve it}
- ...

## Suggested approach (optional)
{When the fix shape is non-obvious, sketch one or more reasonable directions. Skip when the outcome alone makes the fix obvious.}

## Open questions for the human (optional)
- {Any decision an implementing agent shouldn't make unilaterally — pricing thresholds, breaking-change policy, naming, scope of follow-up.}
\\\`\\\`\\\`\\\`

</details>
\`\`\`

**Heading severity emoji** — every \`### \` heading carries one:

- 🚨 critical — blocks merge (data loss, security, broken core flow)
- ⚠️ important — must address before merging (regression, missing validation, incorrect behavior)
- ℹ️ informational — surfaced for awareness; mergeable as-is

**Visible problem write-up rules:**

- **No asks, no suggested fixes** in the visible part. The visible portion describes the problem; the technical-details block describes the fix shape and any open questions. The exception: a fix so self-evident that NOT stating it would be weird (e.g. "the typo is missing an 'r'") — in that case, fold it into the problem statement and skip the suggested-approach block in technical details too.
- **Never two successive plain paragraphs.** Every transition between block-level elements must alternate prose with structure: paragraph → bullet list → paragraph; paragraph → code fence → bullet list; paragraph → table → paragraph. Two consecutive paragraphs in a row create a wall of text that's impossible to digest. If you catch yourself writing one, find a way to split it: pull a list out of it, drop a 2-3 line code fence between them, or merge them into a single tighter paragraph.
- **Per-paragraph budget:** ~3 sentences max. Past that, you're explaining where you should be structuring.
- **Identifier discipline still applies** in the visible part. Lead with behavior in plain English; name an identifier only when it's the subject of the concern or a public surface a reader would recognize. The technical-details block is where dense identifier references belong.

**Technical-details block rules:**

- Wrapped in a 4-backtick markdown fence (\`\\\`\\\`\\\`\\\`markdown ... \\\`\\\`\\\`\\\`\`) so it's visually distinct, one-click copyable, and can contain its own 3-backtick code fences without escape gymnastics. The contents are agent-readable — a fix-agent will pull the body down and use this block as the brief.
- File paths and \`file:line\` refs are encouraged (and necessary) — the next agent uses these to navigate. Identifier density is fine here.
- Slightly more verbose than the absolute minimum is OK when it materially helps the next agent: a small code snippet showing the symptom, a short table of mismatched key/column pairs, a one-paragraph "why CI doesn't catch it" note. Skip massive regression-test scaffolding or full route rewrites — the implementing agent writes those.
- Use the four standard sections (\`Affected sites\`, \`Required outcome\`, optional \`Suggested approach\`, optional \`Open questions for the human\`). Skip the optional sections when they wouldn't add anything.

## 3. \`### ℹ️ Nitpicks\` (optional, last section)

Only when there are nits that for some reason can't be inlined. Filepaths in nit text are fine — these are simple enough that a human or agent reads once and acts. No technical-details block.

\`\`\`
### ℹ️ Nitpicks

- {nit, with file path inline if useful, ≤ ~200 chars}
- ...
\`\`\`

## Body-wide rules

- **Inline-vs-body discipline (repeated for emphasis):** anything that anchors to a single line goes inline, not in the body. The body is for cross-cutting concerns only.
- **No \`### Issues found\` heading** above the issue sections — each \`### \` heading IS the issue.
- **Severity emoji on every \`### \` heading** (🚨 / ⚠️ / ℹ️). No emoji on the preamble lead-in or anywhere else.
- **GitHub block-level rendering**: GitHub's markdown parser requires a blank line between ALL block-level elements (HTML tags like \`<br/>\`, \`<sub>\`, \`<details>\`, \`<b>\` and markdown syntax like headings, lists, blockquotes, code fences, paragraphs). Without a blank line, GitHub treats following content as a continuation of the HTML block and renders markdown syntax as literal text. ALWAYS separate block-level elements with a blank line.
- **Backtick-wrap** every variable, identifier, or file name when you mention one (in either visible or technical-details portions).
- **Don't repeat diff content**, don't include raw \`+123 / -45\` stats, don't include a changelog section, don't use horizontal rules (\`---\`).
- **Pull file/commit counts from \`checkout_pr\` metadata** — never count manually.
- **Legacy headings REMOVED.** Do not use \`### Key changes\`, \`### Issues found\`, \`<b>TL;DR</b>\`, or \`<sub><b>Summary</b>\`. The new structure subsumes them.`;

export function computeModes(agentId: AgentId): Mode[] {
  const t = (toolName: string) => formatMcpToolRef(agentId, toolName);
  return [
    {
      name: "Build",
      description:
        "Implement, build, create, or develop code changes; make specific changes to files or features; execute a plan; or handle tasks with specific implementation details",
      prompt: `### Checklist

1. **task list**: create your task list for this run as your first action.

2. **plan** (optional, for complex tasks): analyze requirements, read AGENTS.md and relevant code, produce a step-by-step implementation plan.

3. **setup**: checkout or create the branch:
   - **PR event, modifying the existing PR**: call \`${t("checkout_pr")}\`
   - **new branch**: use \`${t("git")}\` to create a branch (\`git checkout -b pullfrog/branch-name\`)

4. **build**: implement changes using your native file and shell tools:
   - follow the plan (if you ran a plan phase)
   - plan your approach before writing code: identify which files need to change, key design decisions, and edge cases. for non-trivial changes, consider whether there's a more elegant approach.
   - run relevant tests/lints before committing

5. **self-review**: judgment call — does YOUR diff warrant a fresh-eyes pass?

   Skip self-review (commit directly) when the diff is **genuinely trivial**:
   - doc typos, comment-only edits, whitespace/format-only, import reordering
   - lockfile or generated-code regeneration, mechanical rename whose only effect is import-path updates (size of diff is irrelevant — read the *shape*, not the line count)
   - low-risk dep patch bump from a trusted source

   Run self-review when the diff has **any behavioral surface, however small**:
   - 1-line changes to SQL operators / comparison logic / regexes / redirects / HTTP methods / response codes
   - any change to money / tax / currency / billing / fee / refund / payout calculations or constants
   - any change to auth / permissions / roles / sessions / tokens / signature verification
   - any change to feature-flag defaults, retry counts, timeouts, rate limits, batch sizes
   - new endpoints, new code paths, new error branches — even small ones
   - mixed diffs (whitespace + a single semantic line) — the semantic line still triggers self-review
   - anything you're uncertain about

   Tie-breaker: when in doubt, run self-review. One false-positive subagent dispatch costs cents; one false-negative shipped bug costs much more. There's no value in dispatching for a typo, but there's also no excuse for skipping on a 1-line change to a billing path.

   Otherwise delegate the \`${REVIEWER_AGENT_NAME}\` subagent to review your diff with fresh eyes against YOUR TASK. The subagent's baked-in system prompt enforces a non-mutative + non-recursive contract: read-only file/search/web tools and read-only MCP queries only; no writes, shell side effects, state-changing MCP calls, or nested subagent dispatch. Enforcement is prose-only — restate the constraint in your dispatch instructions and do not relax it.

   Provide the subagent with YOUR TASK, the output of \`git diff\`, and a tight summary (not raw output) of any lint/typecheck/test failures you fixed during build — what broke, root cause, the fix — so it can check that fixes addressed root causes rather than suppressed symptoms; say "no build-phase failures" if the build path was clean. Instruct it to flag bugs, logic errors, missing edge cases, gaps between request and diff, and unintended changes.

   Delegation + research discipline (distilled from \`/anneal\` canonical — these are codified learnings from many review rounds, not theoretical best practices):
   - Do NOT summarize what you implemented — that biases the subagent toward validating the shape of your solution rather than questioning it.
   - Do NOT curate a reading list of files. Let the subagent discover scope from the diff and codebase.
   - Do NOT pre-shape output with a severity / category schema. That leaks your hypotheses; severity is your call during evaluation.
   - Do NOT defect-hunt the diff yourself in parallel with the subagent. Your role is dispatch + evaluation; doing the review yourself reintroduces the implementation bias the subagent is meant to mitigate.
   - For diffs that rely on third-party API contracts, SDK semantics, framework directives, or DB engine specifics, instruct the subagent to verify load-bearing claims via web search and quote source URLs rather than trust training data — this is the single most common review-quality failure mode.

   Be **discerning** about what comes back. The reviewer is an AI subagent and is fallible — treat every finding as a hypothesis, not a directive, and **verify each one yourself** against the diff and the code before deciding whether to apply. You are searching for a solution that is **complete, minimal, and elegant** — you may need to think hard to find it. Do not over-engineer, do not be over-defensive, **do not write AI slop**. Reviewers bias toward *recommending additions*, and that bias has a recognizable slop texture: defensive checks for cases that cannot happen, extra logging, new abstractions used once, comments restating code, tests asserting tautologies, "just-in-case" guards, error handlers for cases the type system already rules out. Reject those. For each surviving finding, ask: would applying it leave the code more sound, correct, AND elegant? Two-out-of-three means look harder for a fix that gets all three before settling. After applying the fixes you accept, re-read your diff and be discerning about what *you just changed*: if any fix turned out to be bloat in context, revert it. Then verify only intended changes are present, no debug artifacts or commented-out code remain, no unrelated files were modified. Commit locally via shell (\`git add . && git commit -m "..."\`).

6. **finalize**:
   - confirm a clean working tree, then push via \`${t("push_branch")}\` (see *SYSTEM* Git rules if this fails — prepush errors are usually the repo's tests/lint, not infra timeouts)
   - create a PR via \`${t("create_pull_request")}\`
   - call \`${t("report_progress")}\` with the PR link or the exact error if push/PR failed

### Notes

For simple, well-defined tasks, skip the plan phase and go straight to build.`,
    },
    {
      name: "AddressReviews",
      description:
        "Address PR review feedback; respond to reviewer comments; make requested changes to an existing PR",
      prompt: `### Checklist

1. **task list**: create your task list for this run as your first action.

2. Checkout the PR branch via \`${t("checkout_pr")}\`.

3. Fetch review comments via \`${t("get_review_comments")}\`.

4. For each comment:
   - understand the feedback
   - **verify the finding yourself** against the actual code before deciding whether to apply — every comment (human or agent) is a hypothesis, not a directive. agent reviewers especially are fallible.
   - you are searching for a solution that is **complete, minimal, and elegant** — you may need to think hard to find it. do not over-engineer, do not be over-defensive, **do not write AI slop**. reviewers bias toward *recommending additions*, and that bias has a recognizable slop texture: defensive checks for impossible cases, extra abstractions used once, comments restating obvious code, tests asserting tautologies, "just-in-case" guards, error handlers for cases the type system already rules out. reject those. evaluate whether applying the finding would leave the code more **sound, correct, AND elegant**; two-out-of-three is a signal to look harder for a fix that gets all three. if a request would add bloat — ceremony without commensurate correctness benefit — push back in your reply rather than mechanically applying it.
   - if the request stands, make the code change using your native tools; otherwise reply explaining why
   - record what was done (or why nothing was done)

5. Quality check:
   - test changes, then review the diff before committing — verify only intended changes are present, no debug artifacts remain, no fix turned out to be bloat in context (revert any that did), and the changes are clean enough that a senior engineer would approve without hesitation
   - commit locally via shell (\`git add . && git commit -m "..."\`)

6. Finalize:
   - confirm a clean working tree, then push via \`${t("push_branch")}\` (same push/prepush guidance as Build mode in *SYSTEM*)
   - reply to each comment **exactly once** using \`${t("reply_to_review_comment")}\` — do not re-emit the same call (the runtime dedupes identical bodies and the second call is wasted)
   - resolve addressed threads via \`${t("resolve_review_thread")}\`
   - call \`${t("report_progress")}\` with a brief summary (or the exact push error if push failed)`,
    },
    // Review and IncrementalReview use a 0-or-2+ lens pattern. The default is
    // 0 lenses (orchestrator handles the review solo). Multi-lens (2+
    // reviewfrog subagents in parallel) only fires for substantive PRs or
    // high-stakes-subsystem touches — and when it fires, ALL lenses must
    // dispatch in a single assistant turn or the parallelism win disappears.
    // We never dispatch exactly one lens: a single lens is just a worse,
    // slower version of doing the work yourself.
    //
    // Build mode self-review is a different problem shape: the orchestrator
    // wrote the code, so bias-mitigation comes from delegating to one
    // fresh-eyes subagent that doesn't share the implementation context. A
    // single subagent there is appropriate; the 0-or-2+ rule applies only to
    // the Review/IncrementalReview lens fan-out where independence between
    // perspectives is what's being purchased.
    //
    // Severity categorization is split across two surfaces: the opening
    // callout (CAUTION/IMPORTANT/ℹ️/✅) sets the review's overall tier, and
    // per-bullet emoji prefixes (🚨/⚠️/ℹ️ in PR_SUMMARY_FORMAT) tag
    // individual points inside summary sections — scoping severity to the
    // specific bullet rather than the whole section keeps a section that
    // mixes a 🚨 and an ℹ️ from being mislabeled by either of them.
    {
      name: "Review",
      description:
        "Review code, PRs, or implementations; provide feedback or suggestions; identify issues; or check code quality, style, and correctness",
      prompt: `### Checklist

1. **task list**: create your task list for this run as your first action.

2. **checkout**: call \`${t("checkout_pr")}\` — this returns PR metadata and a \`diffPath\`. read the diff TOC end-to-end and treat its file line ranges as your coverage checklist.

3. **triage**: orient yourself on the PR — identify *what kind of thing this is* (domain it touches, seams it crosses, external contracts it depends on, user-facing surfaces it changes). pull as much context as you need to render a confident, well-grounded review: read related files, grep for callers of changed symbols, check tests that exercise the touched paths, fetch related GitHub state. **you are the synthesizer** — never delegate understanding to subagents.

   if the PR is **genuinely trivial**, skip the fan-out entirely and submit a \`No new issues found.\` review per step 7.

   "Genuinely trivial" (skip):
   - single-word doc typo, whitespace/format-only, comment-only across any number of files
   - lockfile or generated-code regeneration (size of diff is irrelevant — read the *shape*)
   - mechanical rename whose only effect is import-path updates
   - low-risk dep patch bump

   "Looks trivial but isn't" (do **NOT** skip — small diff, big blast radius):
   - any 1-line change to SQL / regex / auth / billing / permission / signature-verification code
   - flipping a feature-flag default, default config value, or retry/timeout constant
   - changing a money/tax/currency/fee constant by any amount
   - changing an HTTP method, redirect URL, response code, or status enum
   - tightening or loosening a comparison operator (\`<\` ↔ \`<=\`, \`==\` ↔ \`!=\`)
   - renaming a public API surface (still trivial in shape, but needs an impact lens)
   - adding a new direct dependency (supply-chain surface)
   - any "typo fix" in user-facing copy that changes meaning ("approved" → "denied")
   - mixed diffs where a semantic 1-liner is buried in whitespace/formatting changes

4. **lens decision — 0 or 2+, NEVER 1**.

   The default is **0 lenses**: handle the review yourself end-to-end. Most PRs land here.

   Dispatch **2+ \`${REVIEWER_AGENT_NAME}\` lenses in parallel** ONLY when ALL of the following are true:
   - the PR is substantive (>5 files changed AND >200 net lines), OR touches a high-stakes subsystem (auth, billing, payments, schema migration, webhooks, secrets, RBAC, multi-tenant isolation, cron/scheduling)
   - you can name 2+ distinct concrete failure modes that warrant independent lenses (one lens per failure mode; orthogonal, not overlapping)
   - parallel-orchestrated independent perspectives meaningfully outperform what you'd find solo

   **NEVER dispatch exactly one lens.** A single lens is just a more expensive version of doing the work yourself with a worse model — it adds wall time and a context-handoff for no orthogonality benefit. Either you have at least two genuinely independent failure-mode hypotheses (dispatch all in one turn), or you don't (do the review yourself).

   When you do go multi-lens, lens framings come in two flavors:
   - **themed lenses** — a perspective applied across the whole diff (correctness, security, user-journey, performance, etc.).
   - **subsystem lenses** — a domain-scoped frame for high-stakes subsystems the PR touches (e.g. "the auth lens", "the billing lens", "the schema-migration lens"). **for high-stakes domains, lead with the subsystem lens rather than the generic themed equivalent** — "billing-subsystem" outperforms "correctness on billing code" because the framing primes the subagent to remember domain-specific failure modes (double-charges, refund races, currency rounding, dispute flows) the generic lens misses.

   starter menu (combine, omit, or invent your own):
   - **correctness & invariants** — bugs, races, error handling, edge cases, state-machine boundaries
   - **impact** — stale references in code/tests/docs/configs/UI after rename/remove
   - **research-validated assumptions** — third-party API contracts, SDK semantics, framework directives, version-gated behavior. **only pick when the PR's correctness depends on the contract behaving a specific way** — not when the API is merely used. The bar is "if the third-party contract differs from what the diff assumes, the PR is incorrect." When dispatched, the subagent must verify load-bearing claims via web search and quote source URLs.
   - **security** — new endpoints, authZ, input validation, secrets handling, replay/CSRF/injection, cross-tenant isolation
   - **user-journey** — UX-touching flows: walk through happy path and failure modes as a user
   - **operational readiness** — observability, alerting, migrations (forward + rollback), feature flags, on-call burden
   - **integration & cross-cutting** — API contracts between modules, backward-compat of public surfaces, multi-service ordering
   - **test integrity** — meaningful coverage for the changed behavior; deterministic; no shared-state pollution
   - **performance** — N+1 queries, hot-path allocation, latency budgets, index coverage
   - **holistic** — does the PR make sense as a whole? symmetric flows (delete for every create, rollback for every migration)?
   - **subsystem lenses** (invent as the PR demands) — auth, billing, payments, schema migration, webhooks, secrets, RBAC, multi-tenant isolation, cron/scheduling, etc.

   The only subagent type is \`${REVIEWER_AGENT_NAME}\` — used for lens judgment work ("is this safe / correct / well-tested?"), runs on a mid-tier model.

5. **fan out (only if step 4 said 2+ lenses)**: dispatch every \`${REVIEWER_AGENT_NAME}\` subagent for this run **IN A SINGLE ASSISTANT TURN, AS MULTIPLE PARALLEL TASK TOOL_USE BLOCKS IN ONE MESSAGE.**

   ⚠️  CRITICAL — PARALLELISM IS THE ONLY REASON LENSES EXIST. ⚠️
   The default tool-call behavior of Claude Code (and most agent runtimes) is **serial dispatch**: emit one Task call, await result, emit next, await, etc. This collapses your fan-out into a sequential review where each lens adds N × (orchestrator-think-time + lens-execution-time) to wall time. **YOU MUST OVERRIDE THIS DEFAULT.** Emit ALL of your Task tool_use blocks in the SAME assistant message, BEFORE you read ANY result from ANY of them. If you find yourself emitting one Task call, then thinking about the result, then emitting another — STOP and re-issue them all together. The whole point of going multi-lens is the wall-clock speedup from parallel execution; serial dispatch defeats it entirely.

   ✅ Right pattern: one assistant turn with N Task tool_use blocks → wait → N results arrive together → aggregate.
   ❌ Wrong pattern: turn 1 = Task(lens A) → turn 2 (after A's result) = Task(lens B) → turn 3 (after B's result) = Task(lens C). This is the failure mode. Do not do this.

   You can also include your own \`read\` / \`grep\` / \`webfetch\` calls in the SAME turn as the parallel \`${REVIEWER_AGENT_NAME}\` dispatches — concurrent context-pulling on the orchestrator side runs in parallel with the lens fan-out and costs zero extra wall time.

   if a subagent errors out, times out, or returns nothing usable, retry once with the same lens; if it still fails, proceed with partial coverage and note the missing lens in the review body — do not skip the fan-out entirely on a single subagent failure. each subagent gets:
   - the diff path / target — reading the diff and the codebase is its job
   - **only one lens** — never a multi-section "review for X, Y, and Z" prompt
   - **a Task \`description\` set to the lens name** (e.g. \`"security"\`, \`"correctness"\`, \`"billing-subsystem"\`) — the harness reads this field to label the subagent's log lines so parallel runs can be told apart in CI output. without it, every subagent shows up as \`subagent#N\`.
   - if the lens touches external contracts, instruct the subagent to verify load-bearing claims via web search rather than trust training data, and to quote source URLs in its reasoning. action runs are non-interactive — there's no human in the loop to catch "I'm pretty sure Stripe does X."
   - ask the subagent to report findings with file paths and NEW line numbers from the diff so you can anchor inline comments without re-reading the entire diff.

   delegation discipline:
   - do NOT summarize the PR for them (biases toward a validation frame)
   - do NOT hand them a curated reading list (let them discover scope)
   - do NOT pre-shape their output with a finding schema
   - do NOT mention the other lenses (independence is the point — overlapping findings are a strong signal)

6. **aggregate & draft**: when the fan-out lands, merge findings; de-dup overlaps (two lenses catching the same issue = higher-confidence signal); trace each finding yourself before accepting it. drop praise, style preferences, speculative/unverified claims, findings about pre-existing code unrelated to the PR (heuristic: if the finding's root cause lives in lines this PR added or modified, it's in scope; otherwise drop unless the PR plausibly introduced or amplified the regression), and anything not actionable. also drop **bloat-shaped findings** — proposed fixes that would add defensive checks for cases that can't happen, abstractions used once, comments restating obvious code, tests asserting tautologies, or "just-in-case" guards. subagents are fallible and bias toward recommending changes; the bar for an actionable inline comment is sound + correct + elegant. recommending a change that improves only one of the three (or worse, degrades elegance to nominally improve correctness) makes the codebase worse, not better.

   for surviving findings, draft inline comments with NEW line numbers from the diff. every comment must be actionable, 2-3 sentences max. use GitHub permalink format for code references. for impact-analysis findings (stale references after rename/remove), report them in the review body ordered by severity (runtime breakage > incorrect docs > stale comments) rather than as inline comments unless they're anchored to a specific line.

7. **submit**: ALWAYS submit exactly one review via \`${t("create_pull_request_review")}\`. Do NOT call \`report_progress\` — the review is the final record and the progress comment will be cleaned up automatically.

   note: the first create_pull_request_review submission may error with a one-time diff-coverage nudge listing unread TOC regions. retry the same call to proceed — optionally after reading the listed ranges. the pre-flight will not block again this session.

   The review body is structured as: \`[optional alert blockquote]\` → \`[PR summary using the default format below]\`. Inline comments are passed via the \`comments\` parameter, not in the body.

   The opening callout is what the author sees first — pick the one that matches what you want them to do. Five tiers, from loudest to friendliest:

   - \`[!CAUTION]\` — large red banner. Reads as "this will break something."
   - \`[!IMPORTANT]\` — large purple banner. Reads as "you need to look at this before merging."
   - \`> ℹ️ ...\` — informational blockquote. Reads as "minor suggestions, nothing blocking."
   - \`> ✅ ...\` — green friendly blockquote. Reads as "no concerns, mergeable."

   Two reinforcing levers: callout intensity (above) and \`approved\` (which gates the footer Fix-button affordance — Fix renders on every non-approving review, so \`approved: true\` suppresses it). Wrapping mergeable feedback in \`[!IMPORTANT]\` trains users to click Fix on reviews that don't need fixing. Pick the tier the author's actual next action justifies.

   - **critical issues** (blocks merge — bugs, security, data loss, broken core flows):
     \`approved: false\`. Body opens with \`> [!CAUTION]\\n> This PR introduces ...\`, followed by the PR summary. Include all inline comments via \`comments\`.
   - **must-address non-critical findings** (real consequences if shipped — incorrect behavior in non-critical paths, missing validation on user input, regressions the author should fix before merge):
     \`approved: false\`. Body opens with \`> [!IMPORTANT]\\n> ...\`, followed by the PR summary. Reserve this tier for findings with concrete fallout — do NOT use \`[!IMPORTANT]\` for nits, style preferences, or "consider also" suggestions. Include all inline comments via \`comments\`.
   - **minor suggestions only** (single-line nits, doc/comment polish, defer-able observations, "rough edges"):
     \`approved: false\`. Body opens with \`> ℹ️ No critical issues — minor suggestions inline.\\n\\n\` followed by the PR summary. Include all inline comments via \`comments\`. Vary the wording after the emoji to fit the review (e.g. "Minor suggestions only.", "Two rough edges worth a look."), but always keep the ℹ️ prefix and keep it short.
   - **informational observations** (mergeable as-is, nothing actionable — e.g. prior feedback addressed cleanly, surfacing a minor stale doc reference, calling out something noteworthy without recommending a change):
     \`approved: true\`. Body opens with \`> ✅ No new issues found.\\n\\n\` followed by the PR summary. Do NOT include inline \`comments\` — the ✅ signals "no action needed", which contradicts an actionable anchor; if a point is concrete enough to anchor to a line, downgrade the whole review to "minor suggestions only" (\`approved: false\`) instead.
   - **no actionable issues**:
     \`approved: true\`. Body opens with \`> ✅ No new issues found.\\n\\n\` followed by the PR summary.

${PR_SUMMARY_FORMAT}`,
    },
    // IncrementalReview shares Review's 0-or-2+ lens pattern but scopes the
    // target to the incremental diff. The "issues must be NEW since the last
    // Pullfrog review" filter lives at aggregation time (step 8), NOT in the
    // subagent prompt — pushing the filter into
    // subagents matches the canonical anneal anti-pattern of "list known
    // pre-existing failures — don't flag these" and suppresses signal on
    // regressions the new commits amplified. The review body is just
    // "Reviewed changes" — a separate "Prior review feedback" checklist
    // would duplicate the rolling PR summary snapshot's record of what
    // earlier runs already addressed and add noise to the user-facing
    // body. Same opening-callout + per-bullet emoji severity split as Review.
    {
      name: "IncrementalReview",
      description:
        "Re-review a PR after new commits are pushed; focus on new changes since the last review",
      prompt: `### Checklist

1. **task list**: create your task list for this run as your first action.

2. **checkout**: call \`${t("checkout_pr")}\` — this returns PR metadata, \`diffPath\` (full diff), and \`incrementalDiffPath\` (changes since last reviewed version, if available). read the diff TOC first and use its line ranges as your coverage checklist.

3. **incremental scope**: if \`incrementalDiffPath\` is present, read it to see what changed since the last review. this is a range-diff that isolates the net changes, filtering out base branch noise. if not present, fall back to reviewing the full PR diff and determine what changed since Pullfrog's most recent review.

4. **prior feedback**: fetch previous reviews via \`${t("list_pull_request_reviews")}\`. for the most recent Pullfrog review, call \`${t("get_review_comments")}\` with the review ID to retrieve specific prior line-level feedback. you'll use this to filter your aggregation in step 8 — anything already flagged in a prior review and not changed by the new commits should not be re-raised. you do NOT need to render this in the review body; the rolling PR summary snapshot is the durable record of what's been addressed.

5. **triage**: orient on the *incremental* changes — domain, seams, external contracts, user-facing surfaces. pull as much context as you need to render a confident review: read related files, grep for callers of changed symbols, check tests that exercise the touched paths. **you are the synthesizer.**

   if the incremental changes are **genuinely trivial**, skip the fan-out entirely and jump to step 10's non-substantive path (do NOT submit a review).

   "Genuinely trivial" (skip): formatting/comment tweaks, import reordering, lockfile regen, mechanical rename of import paths, whitespace-only.
   "Looks trivial but isn't" (do NOT skip — same anti-patterns as Review mode): 1-line changes to SQL/regex/auth/billing/permissions/signature-verification code; flipping feature-flag defaults or retry/timeout constants; money/tax/HTTP-method/redirect changes; tightening or loosening a comparison operator; mixed diffs with a semantic line buried in formatting.
   When unsure, treat as non-trivial.

6. **lens decision — 0 or 2+, NEVER 1**.

   The default is **0 lenses**: handle the re-review yourself end-to-end. Most incremental reviews land here — especially thread-reply re-reviews where the user is asking "did you address X?" rather than "review the diff again."

   Dispatch **2+ \`${REVIEWER_AGENT_NAME}\` lenses in parallel** ONLY when ALL of the following are true:
   - the incremental changes are substantive (>5 files changed AND >200 net new lines), OR touch a high-stakes subsystem (auth, billing, payments, schema migration, webhooks, secrets, RBAC, multi-tenant isolation, cron/scheduling)
   - you can name 2+ distinct concrete failure modes the new commits plausibly introduce that warrant independent lenses
   - parallel-orchestrated independent perspectives meaningfully outperform what you'd find solo

   **NEVER dispatch exactly one lens.** Single-lens dispatch adds wall time and cost for no orthogonality benefit. Either go multi-lens (≥2 in parallel) or do the re-review yourself.

   Lens framing follows Review mode: themed lenses (correctness, security, etc.) and subsystem lenses (auth, billing, schema-migration, etc.) — for high-stakes domains lead with the subsystem lens.

7. **fan out (only if step 6 said 2+ lenses)**: dispatch every \`${REVIEWER_AGENT_NAME}\` subagent for this run **IN A SINGLE ASSISTANT TURN, AS MULTIPLE PARALLEL TASK TOOL_USE BLOCKS IN ONE MESSAGE.**

   ⚠️  CRITICAL — PARALLELISM IS THE ONLY REASON LENSES EXIST. ⚠️
   Default tool-call behavior is **serial dispatch**: emit one Task call, await result, emit next, await, etc. This collapses your fan-out into a sequential review where each lens adds N × (orchestrator-think-time + lens-execution-time) to wall time. **YOU MUST OVERRIDE THIS DEFAULT.** Emit ALL of your Task tool_use blocks in the SAME assistant message, BEFORE you read ANY result from ANY of them.

   ✅ Right pattern: one assistant turn with N Task tool_use blocks → wait → N results arrive together → aggregate.
   ❌ Wrong pattern: turn 1 = Task(lens A) → turn 2 (after A's result) = Task(lens B). This is the failure mode.

   You can also include your own \`read\` / \`grep\` / \`webfetch\` calls in the SAME turn as the parallel \`${REVIEWER_AGENT_NAME}\` dispatches.

   if a subagent errors out, times out, or returns nothing usable, retry once with the same lens; if it still fails, proceed with partial coverage and note the missing lens in the review body. each subagent gets:
   - the diff scope (incremental diff path if available, full diff otherwise). do NOT tell them to skip pre-existing issues — that suppresses regressions the new commits amplified; the "issues must be NEW" filter lives at aggregation time (step 8), not in the subagent prompt
   - **only one lens** — never a multi-section "review for X, Y, and Z" prompt
   - **a Task \`description\` set to the lens name** — the harness reads this field to label log lines so parallel runs can be told apart.
   - if the lens touches external contracts, instruct the subagent to verify load-bearing claims via web search and quote source URLs.
   - ask the subagent to report findings with file paths and NEW line numbers from the full PR diff so you can anchor inline comments.

   delegation discipline:
   - do NOT summarize the changes for them (biases toward validation frame)
   - do NOT hand them a curated reading list (let them discover scope)
   - do NOT pre-shape their output with a finding schema
   - do NOT mention the other lenses (independence is the point)

8. **aggregate, draft, self-critique**: merge findings (yours + any subagent output if you went multi-lens); de-dup overlaps; trace each finding yourself. drop praise, style preferences, speculative/unverified claims, findings about pre-existing code unrelated to the new commits, anything not actionable, and anything that re-states prior review feedback (heuristic: if the finding's root cause lives in lines the *new commits* added or modified, it's in scope; otherwise drop). also drop **bloat-shaped findings** — proposed fixes that would add defensive checks for cases that can't happen, abstractions used once, comments restating obvious code, tests asserting tautologies, or "just-in-case" guards. subagents are fallible and bias toward recommending changes; the bar for an actionable inline comment is sound + correct + elegant. recommending a change that improves only one of the three (or degrades elegance to nominally improve correctness) makes the codebase worse, not better. To compute "lines the new commits added or modified": if \`incrementalDiffPath\` from step 2 is present, use it directly. Otherwise, take the prior Pullfrog review's \`commit_id\` (returned alongside each entry from \`${t("list_pull_request_reviews")}\` in step 4) and run \`git diff <prior-review-sha>..HEAD\` to isolate the lines added since that review. draft inline comments with NEW line numbers from the full PR diff — every comment must be actionable, 2-3 sentences max.

9. **build the review body** — a single "Reviewed changes" section: summarize at the logical-change level, not per-file. each bullet starts with a past-tense verb (e.g. \`- Extracted shared CLI runtime into a single module\`, \`- Renamed package to pullfrog\`). avoid file paths unless they add clarity. if the changes can be described in one sentence, use one sentence — no bullets needed. do NOT include a separate "Prior review feedback" checklist; that's tracked in the rolling PR summary snapshot for the next agent run, and surfacing it in the user-facing body is noise (changes that addressed prior feedback are already covered by the Reviewed-changes bullets). in some cases you may receive a complete diff for the whole pull request instead of an incremental one — when this happens, you will need to determine what changes have happened since Pullfrog's most recent review.

10. Submit — every run must end with EXACTLY ONE of \`${t("create_pull_request_review")}\` (substantive review) or \`${t("report_progress")}\` (no-review acknowledgement). do NOT call \`create_issue_comment\` for review output.

   Same callout ladder as Review mode — \`[!CAUTION]\` (red, "will break") → \`[!IMPORTANT]\` (purple, "must address before merging") → \`> ℹ️ ...\` (informational, "minor suggestions only") → \`> ✅ ...\` (green friendly, "no concerns"). Same Fix-button lever: the footer renders a Fix button on every non-approving review, so \`approved: true\` suppresses it. Wrapping mergeable feedback in \`[!IMPORTANT]\` trains users to click Fix on reviews that don't need fixing — pick the tier the author's actual next action justifies.

   Follow these rules:
   - note: the first create_pull_request_review submission may error with a one-time diff-coverage nudge listing unread TOC regions. retry the same call to proceed — optionally after reading the listed ranges. the pre-flight will not block again this session.
   - IF NO NEW ISSUES, NON-SUBSTANTIVE CHANGES ONLY (trivial formatting, import reordering, comment tweaks): do NOT submit a review. Instead call \`${t("report_progress")}\` with a 1-2 sentence note explaining no review was warranted (e.g. "No new issues. Changes since last review are formatting-only."). this leaves a visible signal that the run completed.
   - ELSE IF NEW CRITICAL ISSUES (blocks merge — bugs, security, data loss, broken core flows): call \`${t("create_pull_request_review")}\` with \`approved: false\`, all comments, and the review body. body opens with \`> [!CAUTION]\\n> This PR introduces ...\`, then the Reviewed-changes summary.
   - ELSE IF NEW MUST-ADDRESS NON-CRITICAL FINDINGS (real consequences if shipped — incorrect behavior, missing validation, regressions the author should fix before merge): call \`${t("create_pull_request_review")}\` with \`approved: false\`, all comments, and the review body. body opens with \`> [!IMPORTANT]\\n> ...\`, then the Reviewed-changes summary. Do NOT use this tier for nits, style preferences, or "consider also" suggestions.
   - ELSE IF NEW MINOR SUGGESTIONS ONLY (single-line nits, doc/comment polish, defer-able observations, "rough edges"): call \`${t("create_pull_request_review")}\` with \`approved: false\`, all comments, and the review body. body opens with \`> ℹ️ No critical issues — minor suggestions inline.\\n\\n\` (vary the wording after ℹ️ to fit the review), then the Reviewed-changes summary.
   - ELSE IF INFORMATIONAL OBSERVATIONS (mergeable as-is, but worth surfacing — e.g. prior feedback addressed cleanly with one minor stale doc reference, or a noteworthy positive observation): call \`${t("create_pull_request_review")}\` with \`approved: true\`, NO inline comments, and the review body. body opens with \`> ✅ No new issues found.\\n\\n\` (or similar friendly green opener), then the Reviewed-changes summary. If a point is concrete enough to anchor to a line, downgrade the whole review to "minor suggestions only" (\`approved: false\`) instead — the ✅ signals "no action needed", which contradicts an actionable anchor.
   - ELSE IF NO NEW ISSUES, SUBSTANTIVE CHANGES (new functionality, behavior changes, or fixes to prior review feedback): call \`${t("create_pull_request_review")}\` to create a PR review. If all previous reviews have been properly addressed and no new issues were discovered, set \`approved: true\`. body opens with \`> ✅ No new issues found.\\n\\nReviewed the following changes:\\n\`, then the Reviewed-changes summary.`,
    },
    {
      name: "Plan",
      description:
        "Create plans, break down tasks, outline steps, analyze requirements, understand scope of work, or provide task breakdowns",
      prompt: `### Checklist

1. **task list**: create your task list for this run as your first action.

2. Analyze the task and gather context:
   - read AGENTS.md and relevant codebase files
   - understand the architecture and constraints

3. Produce a structured, actionable plan with clear milestones.

4. Call \`${t("report_progress")}\` with the plan.`,
    },
    {
      name: "Fix",
      description:
        "Fix CI failures; debug failing tests or builds; investigate and resolve check suite failures",
      prompt: `### Checklist

1. **task list**: create your task list for this run as your first action.

2. Checkout the PR branch via \`${t("checkout_pr")}\`.

3. Fetch check suite logs via \`${t("get_check_suite_logs")}\`.

4. **CRITICAL**: verify the failure was INTRODUCED BY THIS PR before fixing. If unrelated, abort and report.

5. Diagnose and fix:
   - read the workflow file, reproduce locally with the EXACT same commands CI runs
   - fix the issue using your native file and shell tools
   - verify the fix by re-running the exact CI command
   - review the diff before committing — verify only the fix is present, no debug artifacts, no unrelated changes. the fix should be clean enough that a senior engineer would approve without hesitation.
   - commit locally via shell (\`git add . && git commit -m "..."\`)

6. Finalize:
   - confirm a clean working tree, then push via \`${t("push_branch")}\` (same push/prepush guidance as Build mode in *SYSTEM*)
   - call \`${t("report_progress")}\` with the diagnosis and fix summary (or the exact push error if push failed)`,
    },
    {
      name: "ResolveConflicts",
      description: "Resolve merge conflicts in a PR branch against the base branch",
      prompt: `### Checklist

1. **task list**: create your task list for this run as your first action.

2. **Setup**:
   - Call \`${t("checkout_pr")}\` to get the PR branch.
   - Call \`${t("get_pull_request")}\` to identify the base branch (e.g., 'main').
   - Call \`${t("git_fetch")}\` to fetch the base branch.

3. **Merge Attempt**:
   - Run \`git merge origin/<base_branch>\` via shell.
   - If it succeeds automatically, confirm a clean working tree, push via \`${t("push_branch")}\` (same push/prepush guidance as Build mode in *SYSTEM*), and call \`${t("report_progress")}\` with a brief success note or the exact push error if push failed — **then stop; do not run steps 4–5.**
   - If it fails (conflicts), resolve them manually (continue to steps 4–5).

4. **Resolve Conflicts**:
   - Run \`git status\` or parse the merge output to find the list of conflicting files.
   - For each conflicting file: read it, find the conflict markers (\`<<<<<<<\`, \`=======\`, \`>>>>>>>\`), understand the code context, and rewrite the file with the correct resolution. Remove all markers.
   - Verify the file syntax is correct after resolution.

5. **Finalize**:
   - Run a final verification (build/test) to ensure the resolution works.
   - \`git add . && git commit -m "resolve merge conflicts"\`
   - confirm a clean working tree, then push via \`${t("push_branch")}\` (same push/prepush guidance as Build mode in *SYSTEM*)
   - Call \`${t("report_progress")}\` with a summary of what was resolved (or the exact push error if push failed)`,
    },
    {
      name: "Task",
      description:
        "General-purpose tasks that don't fit other modes: answering questions, adding comments, labeling, running ad-hoc commands, or any direct request",
      prompt: `### Checklist

1. **task list**: create your task list for this run as your first action.

2. Analyze the task. For simple operations (labeling, commenting, answering questions, running a single command), handle directly.

3. For substantial work — code changes across multiple files, multi-step investigations:
   - plan your approach before starting
   - use native file and shell tools for local operations
   - use ${pullfrogMcpName} MCP tools for GitHub/git operations
   - if code changes are needed: review your own diff before committing — verify only intended changes are present, no debug artifacts remain, and the changes are clean enough that a senior engineer would approve without hesitation

4. Finalize:
   - if code changes were made, push to a pull request (new or existing) using \`${t("push_branch")}\` and \`${t("create_pull_request")}\` as needed. \`git status\` must be clean before you finish (see *SYSTEM* Git rules if push fails).
   - call \`${t("report_progress")}\` once with results — include exact tool errors if push or PR creation failed
   - if the task involved labeling, commenting, or other GitHub operations, perform those directly`,
    },
  ];
}

// static export for UI display — uses opencode format as the readable default
export const modes: Mode[] = computeModes("opencode");

/**
 * modes that legitimately never modify the working tree. used by the post-run
 * dirty-tree gate to suppress the "commit and push" nudge — those modes
 * complete by submitting a review (`Review` / `IncrementalReview`) or by
 * posting a Plan comment (`Plan`), not by touching files. any leftover in the
 * tree at end-of-run is incidental tool noise (e.g. a `node_modules/` from a
 * stray install attempt) on an ephemeral worktree; nudging the agent to
 * commit it would produce a spurious PR.
 */
export const NON_COMMITTING_MODES: ReadonlySet<string> = new Set([
  "Review",
  "IncrementalReview",
  "Plan",
]);