geneseed 4.10.0 → 4.14.0

package/src/agents/devops-engineer.md

@@ -34,12 +34,4 @@ Builds CI/CD pipelines and delivery automation for Spring Boot applications.
 ## Output contract
 - Files changed, the stage list and what was validated.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
-
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/docker-expert.md

@@ -34,12 +34,4 @@ Builds small, safe container images for Spring Boot 4 applications.
 ## Output contract
 - Files changed, the build command and whether the image built and started.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
-
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/docs.md

@@ -49,11 +49,4 @@ independently.)*
 - **Receives:** the analyst's brief describing what changed and what needs documenting.
 - **Delivers:** the doc files written/updated, ready for the parent's merge decision.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/empiricist.md

@@ -29,11 +29,4 @@
 - A claims ledger: each load-bearing claim → evidenced? → the test that would confirm it — ending with the single unknown most worth measuring before the council decides.
 - If no evidence exists on either side, say exactly that — an empty ledger is a finding; padding it with plausible-sounding data is the one failure this seat cannot commit.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/explorer.md

@@ -47,11 +47,4 @@ independently.)*
 - **Delivers:** a distilled brief the next seat (developer/docs) can act on
   directly — the scope, the relevant files, and any constraint already found.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/framer.md

@@ -26,11 +26,4 @@
 - The framing read: the real problem in one line, whether the motion fits it, the reframe if the framing is wrong, and the question the council should be debating instead.
 - If the framing holds up under attack, say so plainly and yield — a reframe invented to justify the seat derails a debate that was already sound.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/historian.md

@@ -41,11 +41,4 @@ independently.)*
 - **Delivers:** the precedent and its lesson, same shape as its own output
   contract above.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/kubernetes-specialist.md

@@ -34,12 +34,4 @@ Deploys Spring Boot 4 applications on Kubernetes.
 ## Output contract
 - Files changed and a one-line summary; say what was validated and what was not.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
-
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/memory-curator.md

@@ -61,11 +61,4 @@
   and report what was planned and what was applied — never a partial merge left
   unreported ({{LAW:surface-failures}}).
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/operator.md

@@ -40,11 +40,4 @@ independently.)*
 - **Delivers:** the operability read and the rollback check, same shape as its
   own output contract above.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/performance.md

@@ -62,11 +62,4 @@ independently.)*
 - **Delivers:** measured findings and a verdict, same shape as its own output
   contract above.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/pragmatist.md

@@ -39,11 +39,4 @@ independently.)*
 - **Delivers:** the cheapest viable path and what to cut, before any code is
   written — same shape as its own output contract above.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/react-engineer.md

@@ -0,0 +1,38 @@
+# {{AGENT}}: react-engineer
+
+> {{DESC_REACT_ENGINEER}}
+
+{{CREDIT:everything-claude-code:skills/react-patterns/SKILL.md}}
+
+Implements React features in a Vite or webpack app, TypeScript or JavaScript, following the react and react-testing skills.
+
+## When to dispatch
+- A React component, page, hook or form is scoped and code must be written, or a Vite or webpack build or Jest run is broken and must be made green.
+
+## When NOT to dispatch
+- The design is unclear — use [architect](architect.md) first. The visual direction is open — the frontend-design skill first.
+- Tests only — use [tester](tester.md). Non-React code — use [developer](developer.md).
+
+## Inputs
+- The scoped task, the target folder or worktree, and the acceptance criteria.
+
+## Allowed tools
+- **Read + write to the assigned worktree.** May run the project's lint, typecheck, test and build scripts.
+  <!-- bash: allow -->
+- Never commits or pushes ({{DOCTRINE:consent-before-push}}).
+
+## Procedure
+0. If `{{DIR_MEMORY}}/agents/<your-name>.md` exists, read it first — your durable lessons from prior dispatches ({{DOCTRINE:persist-insight}}).
+1. Read `package.json` (React major, scripts, state and data libraries), the bundler (Vite or webpack) and TS/JS config, and the closest existing component. Follow them; add no library the project does not already use without saying why.
+2. Keep the folder's language: `.tsx` with typed props and no `any`, or `.jsx`.
+3. Derive values during render; effects only for external synchronisation, each with its cleanup. Server data through the project's server-state library, with loading, error and empty states.
+4. Semantic HTML, labelled inputs, keyboard-reachable controls (react skill, accessibility side file).
+5. Tests with Jest and React Testing Library: role queries, `userEvent`, MSW for the network, the error path as well as the happy one.
+6. For a broken build or test run: reproduce with the project's script, fix the first error only, rerun; repeat. Never `@ts-ignore`, `any` or a disabled lint rule to get green. Stop and report after three failed attempts at one error.
+7. Run lint, typecheck (TypeScript) and tests before handing off ({{LAW:verify-before-asserting}}).
+
+## Output contract
+- Files changed, a one-line summary, and each command run with its result.
+- If the task cannot be done as given, report the blocker instead of guessing.
+
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/react-reviewer.md

@@ -0,0 +1,42 @@
+# {{AGENT}}: react-reviewer
+
+> {{DESC_REACT_REVIEWER}}
+
+{{CREDIT:everything-claude-code:agents/react-reviewer.md}}
+
+React-specific review only. Generic correctness, naming and structure belong to [reviewer](reviewer.md); defer to it and do not repeat its checks.
+
+## When to dispatch
+- A change touches `.tsx` / `.jsx` files, hooks or React tests and needs a framework-aware pass.
+
+## When NOT to dispatch
+- No React code in the diff, or a generic review — use [reviewer](reviewer.md).
+- A wider security question (auth design, the API behind the UI) — use [security](security.md).
+
+## Inputs
+- The base..head range (or the diff) and the task it was meant to satisfy.
+
+## Allowed tools
+- **Read-only.** May read code, run the lint, typecheck and test scripts, inspect history.
+- Does not edit code; it reports. Never moves HEAD, the index or the working tree.
+<!-- bash: allow -->
+
+## Procedure
+0. If `{{DIR_MEMORY}}/agents/<your-name>.md` exists, read it first — your durable lessons from prior dispatches ({{DOCTRINE:persist-insight}}).
+1. Read the changed files and `package.json` for the React major. Run lint and, for TypeScript, the typecheck; a missing `react-hooks/rules-of-hooks` or `react-hooks/exhaustive-deps` rule is itself a finding.
+2. Check, in this order:
+   - Security: `dangerouslySetInnerHTML` without a sanitiser at the call site; a user-supplied URL in `href`/`src` without a scheme check; a secret in a bundler-inlined env value (`import.meta.env.VITE_*`, webpack `DefinePlugin`); a session token in `localStorage`.
+   - Hooks: a hook behind a condition, loop or early return; a missing effect dependency, or an `exhaustive-deps` disable without a reason; an effect without cleanup; a stale closure in a timer or async handler.
+   - State: mutation before `setState`; state duplicated or derived through an effect; an effect chain setting state that triggers another effect; `key={index}` on a list that changes.
+   - Data: `useEffect` + `fetch` where the project has a server-state library; a request with no loading, error or empty state; a fetch without abort.
+   - Accessibility: `<div onClick>`, an unlabelled input, `<img>` without `alt`, an error shown by colour alone, a modal that does not manage focus.
+   - Types (TypeScript): `any`, a non-null `!` or an `as` cast that hides a real case, untyped props.
+   - Tests: queries by test id or `querySelector` where a role query works, `fireEvent` where `userEvent` fits, sleeps, component snapshots, the error path untested.
+   - Performance, only with evidence: memoisation with no measured gain, an inline object defeating a `React.memo` child, a long list not virtualised.
+3. Verify by running the tests rather than assuming ({{LAW:verify-before-asserting}}).
+
+## Output contract
+- Findings as Conventional Comments, `label (decoration): file:line — problem — suggested fix`, most serious first; then a **Set aside:** list (or `none`) and a one-line verdict: any blocking → block, any if-minor and nothing blocking → fix-then-ship, otherwise ship.
+- Name any check that was impossible instead of issuing a verdict on partial evidence.
+
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/researcher.md

@@ -39,8 +39,10 @@
    and one empty result is not absence ({{LAW:verify-before-asserting}}).
 3. Open the most promising sources and extract only the relevant slice, never
    the whole page. Prefer primary and recent sources.
-4. Cross-check every material claim against at least two independent sources.
-   A single-source or unsourced claim stays marked unverified.
+4. Cross-check every material claim: one read in the source that owns it (spec, source
+   code, official docs, at the version in question) stands on that source; any other needs
+   two independent sources.
+   Any other single-source or unsourced claim stays marked unverified.
 5. Note recency — flag anything that may be out of date.
 
 ## Output contract
@@ -63,11 +65,4 @@ independently.)*
 - **Delivers:** the sourced findings list, same shape as its own output
   contract above.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/reviewer.md

@@ -55,11 +55,4 @@ independently.)*
 - **Delivers:** the findings list and ship/fix-then-ship/block verdict, same
   shape as its own output contract above.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/security.md

@@ -53,11 +53,4 @@ independently.)*
 - **Delivers:** severity-ranked findings and a ship/fix-required verdict, same
   shape as its own output contract above.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/skeptic.md

@@ -39,11 +39,4 @@ independently.)*
   contract above. Never mechanical proof — this seat runs no commands, so the
   suite and lint results come from the reviewer or tester, not from here.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/spring-boot-engineer.md

@@ -37,12 +37,4 @@ Implements Spring Boot 4 / Java 21+ features, following the spring-boot, spring-
 - Files changed, a one-line summary, and the test command with its result.
 - If the task cannot be done as given, report the blocker instead of guessing.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
-
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/spring-reviewer.md

@@ -38,12 +38,4 @@ Spring-specific review only. Generic correctness, naming and structure belong to
 - Findings as Conventional Comments, `label (decoration): file:line — problem — suggested fix`, most serious first; then a **Set aside:** list (or `none`) and a one-line verdict: any blocking → block, any if-minor and nothing blocking → fix-then-ship, otherwise ship.
 - Name any check that was impossible instead of issuing a verdict on partial evidence.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
-
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/spring-security.md

@@ -39,12 +39,4 @@ Spring Security 7 audit only. The generic audit (injection, secrets, dependencie
 - Findings by severity (critical, high, medium, low), each with `file:line`, the risk and a fix; then a one-line verdict.
 - Name any check that was impossible instead of issuing a verdict on partial evidence.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
-
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/steward.md

@@ -26,11 +26,4 @@
 - A long-term verdict: the structural benefits and risks, the debt this incurs or retires, how reversible it is, and a one-line keep-it-healthy recommendation with the trade-off it accepts.
 - If the change is too small or too reversible to carry long-term stakes, say so and stand down — inflating stakes to justify the seat is its own kind of debt.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/tester.md

@@ -48,11 +48,4 @@ independently.)*
   On failure, findings go back to the developer seat; when the fresh
   developer fails too, the pipeline stops and reports instead of looping again.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/user-advocate.md

@@ -39,11 +39,4 @@ independently.)*
 - **Delivers:** the consumer's verdict, same shape as its own output contract
   above.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/agents/visionary.md

@@ -26,11 +26,4 @@
 - The bold case: the transformative upside in a line, what the motion could become if pushed, the capability it would unlock, and the single bet it depends on. No timid hedging.
 - If the motion genuinely has no bigger version — the ambitious form is just the same thing, larger — say so; a forced 10× pitch is timidity wearing a costume.
 
-## Self-improvement
-
-If this spec misled you — an input you needed but were not given, a boundary
-that proved wrong, a step you could not execute — end your report with one line:
-`spec-feedback: <what failed — the one-line fix>`. Omit it when there is no
-friction. The caller weighs the feedback, folds a real flaw back into this file
-with the user's assent, and records it to {{MEMORY}} only if it clears
-{{DOCTRINE:persist-insight}}'s bar — most reports carry no feedback at all.
+<!-- INCLUDE: agents/_self-improvement.md -->

package/src/kits/react/rules.md

@@ -0,0 +1,13 @@
+### React kit
+
+Applies to React 18 or 19 apps built with Vite or webpack, in TypeScript or JavaScript, tested with Jest and React Testing Library. Read `package.json` first and use only the APIs its React major provides.
+
+- Function components and hooks only. No new class components, except an error boundary where the project has no library for it.
+- Hooks at the top level only. Never disable `react-hooks/rules-of-hooks` or `react-hooks/exhaustive-deps` to make a warning go away.
+- Derive values during render; never copy props or state into state through an effect. Every effect that subscribes cleans up.
+- Server data goes through the project's server-state library, with loading, error and empty states.
+- Keep each folder's language: TypeScript stays typed, with no `any` or `@ts-ignore` to get green; JavaScript stays JavaScript.
+- Env values the bundler inlines (`import.meta.env.VITE_*`, webpack `DefinePlugin`) are public. No secret there, and no session token in `localStorage`.
+- Interactive elements are native (`<button>`, `<a href>`, `<input>` with a label) and keyboard-reachable.
+- Tests query by role or label and act with `userEvent`; the network is mocked with MSW, not by mocking the fetch client.
+- Match the project's existing styling, state library and CI before introducing another.

package/src/kits/react.json

@@ -0,0 +1,18 @@
+{
+  "description": "React on Vite or webpack, in TypeScript or JavaScript, tested with Jest and Testing Library: components, hooks, state, data fetching, accessibility, build and test failures, review.",
+  "skills": ["react", "react-testing"],
+  "agents": ["react-engineer", "react-reviewer"],
+  "rules": "kits/react/rules.md",
+  "rulesWordBudget": 300,
+  "banned": ["ReactDOM.render(", "findDOMNode", "enzyme", "react-test-renderer", "@testing-library/react-hooks"],
+  "catalogue": {
+    "skills": [
+      "| [react]({{DIR_SKILLS}}/react.md) | building or changing React components, hooks, state or data fetching in a Vite or webpack app; bundler config, client-side routing, env variables and build failures; accessibility |",
+      "| [react-testing]({{DIR_SKILLS}}/react-testing.md) | writing or fixing React tests with Jest and Testing Library, MSW network mocks, jest-axe; making Jest run against a Vite or webpack project |"
+    ],
+    "agents": [
+      "| [react-engineer]({{DIR_AGENTS}}/react-engineer.md) | a React component, page, hook or form must be implemented, or a broken Vite or webpack build or Jest run made green |",
+      "| [react-reviewer]({{DIR_AGENTS}}/react-reviewer.md) | a React change needs a framework-aware review: hooks, state, data fetching, accessibility, React security, tests |"
+    ]
+  }
+}

package/src/licenses/everything-claude-code.txt

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Affaan Mustafa
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

package/src/skills/_self-improvement.md

@@ -1,10 +1,23 @@
 ## Self-improvement
 
 <!-- LEAN:begin -->
+This {{SKILL}} keeps its own notes in `{{DIR_NOTEBOOK}}/skill-notes/<name>.md`,
+`<name>` being this {{SKILL}}'s name. When that file exists, read it before the
+first step — where it disagrees with this file, the notes win. They are yours to
+write, and no rebuild touches them.
+
 Close each run with one beat of reflection on the {{SKILL}} itself:
 - A step misled, a needed step was missing, or the trigger fired wrongly — that
-  is a flaw in this file. Propose the exact edit (trigger, procedure, or
-  done-when) and apply it with the user's assent ({{LAW:one-intent-one-act}}).
+  is a flaw in this file. Write the exact correction (trigger, procedure, or
+  done-when) into the notes file, with the user's assent ({{LAW:one-intent-one-act}}), as one line:
+  `- [skill|general] <correction> — seen: <what happened, against what you expected> (<date>)`.
+  `seen:` is the evidence the correction is later judged by; `general` marks one
+  that would fix other {{SKILLS}} too. Never
+  edit this file: it is a rendered copy, a rebuild overwrites it, and the doctor
+  reports the edit as drift. `geneseed diff` carries the notes back to the source.
+- Keep the notes sharp: amend or delete a line before adding one — a line with no
+  `seen:`, or whose `seen:` no longer happens, goes first — and prune past about
+  twenty lines.
 - A lesson that is *not* a flaw in this file goes to {{MEMORY}} only if it
   clears {{DOCTRINE:persist-insight}}'s bar: it would change how a future session behaves, and a
   fresh read of the repo would not re-derive it. Update an existing memory over
@@ -13,7 +26,11 @@ Close each run with one beat of reflection on the {{SKILL}} itself:
   runs end here.
 <!-- LEAN:else -->
 One beat of reflection on this {{SKILL}}: a step that misled, a missing step, or a
-wrong trigger is a flaw in this file — propose the exact edit and apply it with the
-user's assent ({{LAW:one-intent-one-act}}). A lesson that is not about this file goes to {{MEMORY}} only
-if it clears {{DOCTRINE:persist-insight}}'s bar. Nothing learned — move on.
+wrong trigger is a flaw in this file — write the exact correction, with the user's
+assent ({{LAW:one-intent-one-act}}), into `{{DIR_NOTEBOOK}}/skill-notes/<name>.md` (`<name>` is this {{SKILL}}'s
+name) as `- [skill|general] <correction> — seen: <what happened vs expected> (<date>)`,
+never into this rendered file, which a rebuild overwrites. Read that notes
+file before the first step when it exists; it wins where it disagrees. Keep it short.
+A lesson that is not about this file goes to {{MEMORY}} only if it clears
+{{DOCTRINE:persist-insight}}'s bar. Nothing learned — move on.
 <!-- LEAN:end -->

package/src/skills/brainstorm.md

@@ -7,6 +7,8 @@
 
 {{CREDIT:mattpocock-skills:skills/productivity/grilling/SKILL.md}}
 
+{{CREDIT:mattpocock-skills:skills/productivity/to-questionnaire/SKILL.md}}
+
 **Trigger:** a new feature, behaviour change, or project arrives without a design, or the
 user says "brainstorm", "interview me" or "grill me". Also when goal, scope or success
 criteria are unstated ("let's design this", "clarify this first", "stress-test this plan"),
@@ -27,8 +29,11 @@ the design is skipped.
    libraries) the user wants to own. Existing work decides by default
    ({{DOCTRINE:respect-conventions}}); a choice you take goes in the ledger as **your
    call**, never as agreed. Stop when each is unambiguous. Facts are yours to find (read
-   the code, dispatch a subagent), never the user's. A decision someone else owns gets a drafted
-   `questionnaire-<slug>.md` for them — ask the user only who and what they need back.
+   the code, dispatch a subagent), never the user's; a pending lookup holds back only the
+   questions that depend on it. A decision someone else owns gets a drafted
+   `questionnaire-<slug>.md` for them — ask the user only who and what they need back. Most
+   important first (async gets one pass), one idea per question with an empty answer stub
+   under it, and say "I don't know" is a useful answer.
 3. **Ledger the decisions.** Name every KEY DECISION the answers imply or leave open —
    direction chosen, trade-offs accepted, load-bearing constraints, assumptions,
    non-goals. Surface each silent assumption as a decision to ratify, not a settled

package/src/skills/codebase-design.md

@@ -16,7 +16,8 @@ a design talk that keeps stumbling over ambiguous terms, or another {{SKILL}} ne
    `GLOSSARY-MAP.md` means several contexts — find the one in play) and the
    area's ADRs in `docs/adr/`; use their terms verbatim in code, tests, prose.
    When the user says how something works, check the code agrees; surface any
-   contradiction.
+   contradiction. An ADR is settled: a proposal that contradicts one says so,
+   and why the friction justifies reopening it — never silently.
 2. Use this vocabulary exactly — **module** (anything with an interface and an
    implementation, any scale), **interface** (all a caller must know: types,
    invariants, ordering, error modes, performance), **depth** (behaviour per
@@ -25,7 +26,10 @@ a design talk that keeps stumbling over ambiguous terms, or another {{SKILL}} ne
    "component", "service", "API", "boundary".
 3. Aim for **deep modules** — much behaviour behind a small interface: callers
    get leverage, maintainers get locality. To find what to deepen, start from
-   `git log` hot spots — what keeps changing.
+   `git log` hot spots — what keeps changing. Design a new or reshaped interface
+   twice before choosing: 2–3 radically different shapes (in parallel subagents
+   where the host runs them), compared on depth, locality and seam placement;
+   recommend one — a hybrid if parts combine — never a menu.
 4. Seams sparingly: the interface is the test surface. Internal seams for the
    module's own tests are fine but never exposed; a test that must reach past
    the interface means the wrong shape. One adapter is a hypothesis; a second

package/src/skills/debug.md

@@ -28,6 +28,6 @@ faster than thrashing; do not trade it away under pressure.
 **CI branch (a red run)** → read [debug/ci.md](debug/ci.md) and follow it.
 
 ## Done when
-- The failure is reproduced by a command that went red on it, root-caused, fixed at the cause, and that command passes with no new breakage; every tagged probe is gone.
+- The failure is reproduced by a command that went red on it, root-caused, fixed at the cause, and that command passes with no new breakage; every tagged probe is gone, and the step-1 repro is deleted or kept as the regression test.
 
 <!-- INCLUDE: skills/_self-improvement.md -->

package/src/skills/develop.md

@@ -33,7 +33,7 @@ case where the seam is only known once the change is made.
      change ({{DOCTRINE:cover-and-verify}}), then break the implementation on purpose once to
      see the test go red before trusting it.
 4. Write the minimum code that makes it pass; add nothing the test doesn't demand.
-5. Run the *affected* tests and read the output ({{LAW:verify-before-asserting}}); on failure fix the
+5. Run the *affected* tests, and the typecheck where the project has one, and read the output ({{LAW:verify-before-asserting}}); on failure fix the
    cause, never mask it ({{LAW:cure-the-cause}}). Dispatch the
    [tester {{AGENT}}](../{{DIR_AGENTS}}/tester.md) for heavier or unfamiliar coverage,
    or when the blast radius is unclear.
@@ -50,7 +50,9 @@ case where the seam is only known once the change is made.
 ## Anti-patterns — name the one you catch
 - **Implementation-coupled** — mocks an internal collaborator, tests a private method,
   or checks through a side channel (querying the database instead of the interface).
-  The tell: it breaks on a refactor that changed no behaviour.
+  The tell: it breaks on a refactor that changed no behaviour. Mock only at system
+  boundaries — external APIs, time, randomness, sometimes the database or filesystem —
+  passed in rather than built inside; never a module you control.
 - **Tautological** — the expected value is recomputed the way the code computes it, so
   the test passes by construction and can never disagree with the code.
 - **Horizontal slicing** — all the tests first, then all the code: the tests pin an

package/src/skills/handoff.md

@@ -11,7 +11,7 @@
 2. Record open decisions, dead ends already ruled out, blockers, and any irreversible changes already made ({{DOCTRINE:plan-before-acting}}) — so the next agent neither re-derives nor repeats them.
 3. Promote any durable fact or correction learned this session into {{MEMORY}} (§5), not just the worklog.
 4. Point to the live artifacts: branch name, changed files, how to run the tests, and where any failure reproduces. Link specs, plans, ADRs, issues and commits by path or URL — never copy them in.
-5. Put a one-line "resume here" pointer at the top so a fresh context picks up in a single read ({{DOCTRINE:context-economy}}), and list the {{SKILL}}s the next agent should load.
+5. Put a one-line "resume here" pointer at the top so a fresh context picks up in a single read ({{DOCTRINE:context-economy}}), and list the {{SKILL}}s the next agent should load. When the user names what the next session is for, write toward that focus.
 6. Redact keys, passwords and PII ({{LAW:sealed-secrets}}).
 7. Verify it: re-read the worklog as a cold reader ({{LAW:verify-before-asserting}}) — goal, progress, next step, and blockers must be answerable from it alone. Fix what isn't.