kodelyth-ecc 2.19.0 → 2.21.0

package/.agent/rules/agent-intent-routing.md

@@ -1,832 +0,0 @@
-# Agent Intent Routing — Auto-Detect User Intent → Right Specialist Agent
-
-This rule teaches the AI to **automatically route the user to the correct specialist agent** based on what they say, write, or feel — without requiring `use <agent-name>` syntax.
-
-When any of the patterns below match the user's message, the AI MUST:
-
-1. **Acknowledge** the routing in one short line: `→ Routing to <agent-name>`
-2. **Behave as that agent** for the rest of that response (apply its persona, methodology, and constraints)
-3. **Suggest the explicit invocation** at the end so the user learns the toolkit: `Tip: next time you can type "use <agent-name>"`
-
-If multiple agents match, pick the **highest priority** below (priority decreases top-to-bottom). If two are equally relevant, name both and ask the user which to use.
-
-**Never silently route.** Always be transparent about which agent is taking over.
-
-> **Companion rule:** [`cost-aware-model-routing`](./cost-aware-model-routing.md) handles the orthogonal question of _which model tier_ to use for the routed task. They run together — agent routing picks the specialist; model routing picks the right tier so the team isn't burning frontier-model tokens on doc typos.
-
----
-
-## Routing v2 — The Eight Dimensions
-
-The signal tables below tell you **which agent** matches. These dimensions tell you **how confidently to route, when to stay silent, how to announce, and how to handle compound intent**. Apply them in order.
-
-### 1. Confidence tiers
-
-| Tier | Trigger | AI behaviour |
-|---|---|---|
-| **High** — 2+ signals from one table, OR 1 signal + emotion, OR explicit `use X` / `@X` | Route immediately with one-line announcement |
-| **Medium** — 1 clean signal, no counter-signals | Route with one-line announcement AND a "Not X? Say so, I'll switch" tail |
-| **Low** — 1 weak signal, or signals from 2+ tables of similar priority | **Do NOT auto-route.** Name the 2 best candidates in one line, ask which fits |
-| **None** — no signal, or explicit anti-routing | Answer directly, no routing announcement |
-
-### 2. Session-state awareness (sticky routing)
-
-Once you have routed to an agent this session, **stay in that agent's voice** for follow-ups on the same thread unless:
-
-- The user's next message contains a **stronger signal** for a different agent
-- The user says "back to normal", "stop routing", "just answer me"
-- The task has clearly concluded (test passes, PR merged, question answered)
-
-**Never re-announce the same routing.** Sticky = quiet. Only announce again when the agent actually changes.
-
-### 3. Anti-routing whitelist
-
-Route to **no one** in these cases (skip Priority 1-10 entirely):
-
-- Message is under 5 words AND has no error/code paste (e.g. "hi", "ok", "thanks", "cool", "yes")
-- Message is a factual question with a one-line answer ("what does `git stash pop` do?")
-- User already invoked with `use X`, `@X`, `run <agent>` — respect explicit choice
-- User said "don't route" / "just answer" / "stop routing" this session
-- User is in the middle of a multi-step interactive workflow the previous agent started
-- The message is a bug report about ECC itself (routing bugs, install bugs) — answer as the maintainer, not as a specialist
-
-### 4. New signal families (v2.4+)
-
-Beyond the 10 priority tiers below, watch for these newer signals — they map to features shipped in ECC 2.x:
-
-| User says | Route to |
-|---|---|
-| "compress this memory file" / "shrink CLAUDE.md" / "make my rules smaller" | `/terse-compress` |
-| "talk shorter" / "be brief" / "less verbose" / "caveman mode" | `/terse` (skill activation) |
-| "who calls X" / "where is X used" / "trace the call chain" / "impact analysis" | `codebase-memory-mcp` via `search_graph` / `trace_path` |
-| "index this project" / "build the graph" / "scan the codebase" | `codebase-memory-mcp` via `index_repository` |
-| "remove ECC" / "uninstall kodelythecc" / "clean this up" | Suggest `kodelythecc uninstall --dry-run` |
-| "how much did I save" / "show my token savings" / "RTK stats" | `kodelythecc dashboard` OR `kodelythecc rtk gain` |
-| "install for another IDE" / "add to Cursor / Windsurf / Antigravity" | `kodelythecc --target <ide>` |
-
-### 5. Compound intent → parallel agents
-
-Some messages match **two priority tables at once**. When they do, fire the parallel command, not both agents sequentially.
-
-| User signals | Fire |
-|---|---|
-| security question + review question | `/security-audit` |
-| bug + been-stuck-for-hours + multi-layer | `/debug-blitz` |
-| architecture + security + a11y (new project) | `/project-launch` |
-| pre-release + full audit | `/pre-release` |
-| refactor + types + tests | `/refactor-sprint` |
-| attacker mindset + multi-vector | `/devil-mode` |
-| new codebase + orientation | `/onboard` |
-| general audit + multiple angles | `/team-review` |
-
-### 6. Announcement style adapts to terse mode
-
-If the user has typed `/terse` this session:
-
-- Drop the `Tip: next time you can type "use <agent>"` line
-- Use a one-token announcement: `→ debug-detective` instead of `→ Routing to debug-detective (stack trace + frustration match)`
-- No trailing decoration
-
-If terse is **off**, use the full form documented in the "Output Format" section at the bottom.
-
-### 7. Cultural + multi-language cues
-
-The user may write in a mixed register (English + native language + technical shorthand). Read the cues:
-
-- Emotional markers in any language route the same way: "arre yaar broken hai" = frustration = `debug-detective`
-- Filler like "bro", "yaar", "man", "please" is not a signal — strip it, read the substance
-- If the user's message language differs from previous turns, respond in the same language they used **this turn**, not their previous one
-- Code identifiers, commands, and error text stay in their original language (never translate `NullPointerException`)
-
-### 8. Evolve integration — routing misses become proposals
-
-When you route (or explicitly do NOT route because the signal was weak) and the outcome is unsatisfying, the ECC evolve system writes this to `~/.kodelythecc/evolve/routing-misses.jsonl` so proposals for new triggers can be surfaced later.
-
-You never write this file yourself. Just be honest in the routing announcement: if the match is weak, say "medium confidence — routing to X, not Y, because Z" so the evolve pipeline has a clean signal.
-
----
-
-## How to Apply This Rule — Semantic Intent, Not Keyword Matching
-
-**Read the intent behind the words, not just the words.**
-
-- If someone says "this thing is broken", they mean `debug-detective` even though they didn't say "bug"
-- If someone says "help me make a todo app", they mean `/project-launch` even though they didn't say "project-launch"
-- If someone pastes a block of code with no context, treat it as a review request → `code-reviewer`
-- If someone pastes a stack trace or error log with no context, treat it as a debug request → `debug-detective`
-- If the user's message expresses **emotion** (frustration, excitement, confusion, worry), that is a routing signal
-
-**The signal tables are examples, not exhaustive lists.** Use them to understand the category of intent, then reason freely about whether the current message fits.
-
-**When in doubt between two agents:** name both options in one line and ask "Which fits your situation better?" — do not silently pick one.
-
----
-
-## Priority 1 — Crisis & Emotional Signals
-
-### `kodelyth-advisor` — User is lost, stuck, overwhelmed
-
-Trigger if the user expresses **uncertainty about direction**, not a specific technical question.
-
-| Signal | Real human phrasing |
-|---|---|
-| Stuck / lost | "I'm stuck", "I'm lost", "I have no idea", "I don't know where to start", "no clue what to do" |
-| Overwhelmed | "this is too much", "I'm overwhelmed", "where do I even begin", "there's so much to think about" |
-| Asking for direction | "what should I do", "help me figure out", "should I X or Y", "I'm not sure which way to go" |
-| Confused about the codebase | "I don't understand this code", "how does this even work", "what is all this doing" |
-| First-time on a problem | "first time doing X", "never done this before", "is there a right way", "I'm new to this" |
-| Seeking reassurance | "am I doing this right?", "is this approach okay?", "what would you do here?" |
-| General help request | "can someone help me with this?", "any advice?", "I need guidance" |
-
-**Counter-signals (do NOT route here):** specific error message, stack trace, named file/function, or a direct technical question with a clear answer.
-
-### `pair-programmer` — User is about to write code
-
-Trigger if the user describes **what they're about to build** before they start.
-
-| Signal | Real human phrasing |
-|---|---|
-| Pre-implementation | "I want to build", "I'm going to add", "let's create", "I need to make a", "let me add", "I'm starting to work on" |
-| Approach question | "how should I implement", "what's the best way to", "should I use X pattern", "which approach is better" |
-| Architecture sketch | "I'm thinking of doing X then Y", "my plan is to", "I was thinking maybe" |
-| Pre-flight check | "before I start", "thinking about", "planning to", "about to write" |
-| Asking how | "how do I build", "how would you do", "what's the pattern for", "is there a standard way" |
-| Single feature | User describes adding ONE feature or function to an existing codebase |
-
----
-
-## Priority 2 — Active Pain (something is broken)
-
-### `debug-detective` — Bug, error, unexpected behavior
-
-| Signal | Real human phrasing |
-|---|---|
-| Direct bug report | "bug", "broken", "crashed", "exception", "stack trace", "traceback", "error" |
-| Frustration | "I've been trying for hours", "can't figure out why", "this won't work", "driving me crazy", "nothing works", "I'm losing my mind" |
-| Unexpected behavior | "why is this X", "shouldn't this Y", "not what I expected", "acting weird", "behaving strangely" |
-| Something is wrong | "something's wrong", "it's broken again", "stopped working", "used to work and now doesn't" |
-| Specific error keywords | `TypeError`, `NullPointerException`, `panic`, `segfault`, `undefined`, `cannot read property`, `500 error`, `404` |
-| Test failures | "test fails", "expected X got Y", "assertion failed", "tests are red", "CI is failing" |
-| Implicit (code paste) | User pastes a stack trace, error log, or failing test output with no other context |
-
-**Always:** ask for the exact error, the minimal repro, and what was changed last.
-
-### `silent-failure-hunter` — Something is wrong but no error
-
-| Signal | Real human phrasing |
-|---|---|
-| Mismatch without error | "no error but wrong result", "data is wrong", "looks fine but isn't working", "getting the wrong output" |
-| Silent state corruption | "X happened but Y didn't fire", "the value is off", "race condition", "something is off" |
-| Swallowed exceptions | "try/catch", "fallback", "default value", "it just returns null", "it silently fails" |
-| Puzzling behavior | "it's doing something unexpected", "the numbers don't add up", "data seems off", "I'm confused by the output" |
-
-### `build-error-resolver` — Build / compile / type errors
-
-| Signal | Real human phrasing |
-|---|---|
-| Build failed | "build failed", "compile error", "won't compile", "cargo build", "npm run build", "build is red", "failing to build" |
-| Type errors | "type error", "TS2322", "type mismatch", "cannot assign", "incompatible types", "TypeScript is yelling at me" |
-| Module/import | "module not found", "cannot resolve", "import error", "can't find the module" |
-| Deploy errors | "Vercel build failed", "Railway is erroring", "can't deploy", "deployment is broken" |
-
-**Language-specific routing:** Go → `go-build-resolver`, Rust → `rust-build-resolver`, Java → `java-build-resolver`, Kotlin → `kotlin-build-resolver`, C++ → `cpp-build-resolver`, Dart/Flutter → `dart-build-resolver`, PyTorch → `pytorch-build-resolver`.
-
----
-
-## Priority 3 — Quality & Review
-
-### `code-reviewer` — User wants feedback on existing code
-
-| Signal | Real human phrasing |
-|---|---|
-| Direct review request | "review this", "code review", "LGTM?", "thoughts?", "feedback", "take a look at this", "eyes on this" |
-| Quality check | "is this good", "is this clean", "anything wrong", "what do you think of this code", "can you check this" |
-| Refactor opportunity | "is there a better way", "smell test", "looks ugly", "feels messy", "is this readable" |
-| Implicit (code paste) | User pastes code with minimal text — treat as a review request |
-| Sanity check | "does this make sense?", "is this the right approach?", "would you write it differently?" |
-| Before merge | "about to merge this", "is this PR-ready?", "pre-merge check" |
-
-**Key rule:** Single-file or single-function review → language reviewer. Full project or multiple files → `/team-review`.
-
-**Language-specific routing (if file extension or language is mentioned):**
-
-| Language | Agent |
-|---|---|
-| TypeScript / JavaScript | `typescript-reviewer` |
-| Python | `python-reviewer` |
-| Go | `go-reviewer` |
-| Rust | `rust-reviewer` |
-| Java / Spring | `java-reviewer` |
-| Kotlin / Android | `kotlin-reviewer` |
-| C++ | `cpp-reviewer` |
-| C# / .NET | `csharp-reviewer` |
-| Flutter / Dart | `flutter-reviewer` |
-| SQL / Postgres / Supabase | `database-reviewer` |
-| Healthcare / EMR / PHI | `healthcare-reviewer` |
-
-### `security-reviewer` — Security concern
-
-| Signal | Real human phrasing |
-|---|---|
-| Direct security question | "is this secure", "vulnerability", "CVE", "audit", "any security issues", "is this safe" |
-| Common attacks | "SQL injection", "XSS", "CSRF", "SSRF", "auth bypass", "RCE", "path traversal", "injection" |
-| Auth & secrets | "auth flow", "JWT", "session", "API key in code", "leaked secret", "exposed credential", "token in code" |
-| Crypto | "encryption", "hashing passwords", "salt", "AES", "RSA", "how to store passwords" |
-| Input handling | "user input", "sanitize", "escape", "validation", "user-submitted data" |
-| Worry signal | "I'm worried this might be insecure", "is there a risk here", "could someone exploit this" |
-
-### Devil-Mode Adversarial Crew (8 specialists — attacker mindset)
-
-These 8 agents look at the same code an attacker would. Route to them when the user shows signs of adversarial thinking, supply-chain worry, AI safety concern, or pre-public-launch paranoia.
-
-#### `prompt-injection-hunter` — AI-feature attack-surface review
-
-| Signal | Real human phrasing |
-|---|---|
-| AI feature audit | "is my AI feature safe?", "can my agent be jailbroken?", "review my prompts", "audit my LLM integration" |
-| Prompt safety | "prompt injection", "indirect injection", "system prompt leak", "tool hijacking", "RAG security" |
-| MCP server review | "is this MCP server safe?", "audit my MCP integration", "agent tool security" |
-
-#### `supply-chain-auditor` — Dependency & install-script attacker review
-
-| Signal | Real human phrasing |
-|---|---|
-| Dep paranoia | "is this package safe?", "could this dep be malicious?", "typosquat", "dependency confusion", "is npm install safe?" |
-| Lockfile concern | "lockfile drift", "should I trust this dep?", "audit my deps for malware", "what's in my node_modules" |
-| Recent CVE worry | "is this CVE in my supply chain", "compromised package alert", "is `<pkg>` malicious now?" |
-
-#### `secret-hunter` — Leaked credential hunt
-
-| Signal | Real human phrasing |
-|---|---|
-| Leak worry | "did I leak any secrets?", "scan for API keys", "any tokens in my code?", "secret scanner", "credentials in git" |
-| Pre-public | "I'm about to make this public, any secrets in history?", "before I go open-source check secrets" |
-| Bundled to client | "are my env vars exposed to the browser?", "is `process.env.X` leaking to the client?" |
-| Rotate event | "rotated my keys, did I miss any?", "post-incident sweep" |
-
-#### `license-violation-finder` — IP / copyleft contamination audit
-
-| Signal | Real human phrasing |
-|---|---|
-| License audit | "license audit", "are my deps GPL?", "license compatibility", "copyleft check", "AGPL contamination" |
-| Going public | "before I open-source, license audit", "any license issues with my deps?" |
-| Due diligence | "M&A license review", "investor due diligence", "what licenses are in my distribution" |
-| Attribution | "do I need to add attribution?", "missing THIRD-PARTY.md", "MIT license requirements" |
-
-#### `jailbreak-tester` — Live AI feature red-team
-
-| Signal | Real human phrasing |
-|---|---|
-| AI red-team | "red-team my AI", "test my chatbot for jailbreaks", "can my LLM be tricked?", "AI safety testing" |
-| Pre-launch AI | "launching an AI feature, is it safe?", "production AI safety check", "test my agent for refusal bypasses" |
-| Overrefusal | "my AI refuses too much", "false refusals", "my chatbot won't answer normal questions" |
-
-#### `code-stealer-detector` — Code provenance audit
-
-| Signal | Real human phrasing |
-|---|---|
-| Provenance | "where did this code come from?", "provenance check", "audit copy-paste origin", "is this Stack Overflow code?" |
-| AI-generated | "audit AI-generated code", "Copilot-written code provenance", "is my LLM-generated code safe to ship?" |
-| Pre-public | "before going open-source, check code origins", "M&A IP audit", "due diligence code review" |
-| Style break | "this code looks pasted", "doesn't match our style", "audit unusual code blocks" |
-
-#### `backdoor-hunter` — Malicious-code pattern detection
-
-| Signal | Real human phrasing |
-|---|---|
-| Malware suspicion | "is this code malicious?", "look for backdoors", "obfuscated code", "what does this eval do?", "scan for hidden payloads" |
-| Post-compromise | "after a security incident, sweep the code", "vendor said they were compromised, audit my install" |
-| Vendored review | "audit this vendored library", "review this third-party code dump", "contractor code review for malice" |
-| Suspicious CI | "weird stuff happening in my CI", "build is doing things it shouldn't", "unexplained network calls" |
-
-#### `chaos-engineer` — Adversarial reliability / fault injection
-
-| Signal | Real human phrasing |
-|---|---|
-| Resilience | "test my system's resilience", "chaos engineering", "fault injection", "what happens if X dies?" |
-| Pre-launch readiness | "is my system production-ready under failure?", "test failure modes before launch" |
-| Hidden assumption | "what hidden assumptions am I making?", "what could break?", "stress my dependencies" |
-| Game day | "schedule a game day", "team chaos exercise", "training for incidents" |
-
-**Counter-signal (for all devil-mode agents):** General quality concern not framed adversarially → use the standard `code-reviewer` / `security-reviewer` / `performance-optimizer`. Route to devil-mode only when the user is explicitly thinking like an attacker or asking pre-public/pre-launch questions.
-
-### `api-guardian` — API contract changes
-
-| Signal | Real human phrasing |
-|---|---|
-| API change | "breaking change", "deprecate", "API version", "backwards compat", "changing the endpoint" |
-| Contract concern | "consumer impact", "client breakage", "field removal", "rename endpoint", "will this break clients" |
-| Spec work | "OpenAPI", "GraphQL schema", "REST design", "idempotent", "API design review" |
-
-### `ux-reviewer` — Frontend / UX / Accessibility
-
-| Signal | Real human phrasing |
-|---|---|
-| UI feedback | "feels off", "looks weird", "UX", "user experience", "interaction", "the UI is confusing" |
-| Accessibility | "a11y", "WCAG", "screen reader", "ARIA", "keyboard nav", "color contrast", "accessible" |
-| Form / interaction | "form validation", "loading state", "error state", "empty state", "the form feels clunky" |
-| Mobile / responsive | "mobile", "responsive", "viewport", "touch target", "doesn't look right on phone" |
-| Usability | "users are confused by", "hard to use", "the flow is broken", "nobody can find the button" |
-
----
-
-## Priority 4 — Performance & Scale
-
-### `incident-commander` — Production is down or degraded
-
-Route here FIRST for any active production incident. This takes priority over `debug-detective` when the incident is live in production.
-
-| Signal | Real human phrasing |
-|---|---|
-| Production down | "production is down", "outage", "site is down", "service unavailable", "prod is broken" |
-| P0 / P1 | "P0", "P1", "incident", "on-call", "pagerduty fired", "alert triggered", "woke up to alerts" |
-| Blast radius | "10% of users affected", "all requests failing", "error rate spiked", "users can't login" |
-| Active degradation | "production is throwing 500s", "latency is through the roof", "database is down", "everything is slow in prod" |
-| Postmortem | "postmortem", "incident review", "blameless review", "what went wrong", "write up the incident" |
-
-**Counter-signals:** development bug (not production), staging environment, local testing — route to `debug-detective` instead.
-
-### `load-tester` — Load, stress, and capacity testing
-
-| Signal | Real human phrasing |
-|---|---|
-| Load test request | "load test", "stress test", "performance test", "capacity test", "can it handle X users" |
-| Tools | "k6", "Locust", "Artillery", "wrk", "Gatling", "hey", "ab test" |
-| Capacity planning | "how many users can we handle", "what's our breaking point", "max RPS", "will it scale" |
-| Pre-launch validation | "will this hold under load", "ready for launch traffic", "scale test", "launch is tomorrow" |
-| Soak test | "soak test", "memory leak under load", "sustained load test", "long-running load" |
-
-**Counter-signal:** "make this code faster" → `performance-optimizer`. Load-tester handles test design, not code optimization.
-
-### `performance-optimizer` — Slowness / bottleneck in code
-
-| Signal | Real human phrasing |
-|---|---|
-| Slow | "slow", "sluggish", "laggy", "takes forever", "timing out", "too slow", "feels sluggish" |
-| Resource | "high CPU", "memory leak", "OOM", "out of memory", "cpu pinned", "memory keeps growing" |
-| Frontend perf | "FCP", "LCP", "INP", "bundle size", "render time", "jank", "the page loads slowly" |
-| Backend perf | "N+1", "slow query", "throughput", "latency", "p99", "the API is slow" |
-| Optimization request | "make this faster", "optimize this", "reduce the load time", "improve performance" |
-
----
-
-## Priority 5 — Planning & Architecture
-
-### `planner` — Feature planning
-
-| Signal | Real human phrasing |
-|---|---|
-| Plan a feature | "plan this", "roadmap", "break down", "sprint plan", "milestones", "break this into tasks" |
-| Decompose | "where do I start with X feature", "tasks for", "work items", "list the steps for" |
-| Scope definition | "what do we need to build for X", "what's involved in adding X" |
-
-### `architect` — System design
-
-| Signal | Real human phrasing |
-|---|---|
-| System-level | "architecture", "system design", "how should the services interact", "what's the overall design" |
-| Choosing tech | "should I use Postgres or Mongo", "monorepo vs polyrepo", "REST vs GraphQL", "which database" |
-| Scaling | "horizontal scaling", "sharding", "queue", "event-driven", "how do I scale this" |
-| Design decision | "what's the right architecture for X", "how should I structure this", "where should this live" |
-
-### `code-architect` — Single-feature blueprint
-
-Trigger when the user wants a **detailed implementation blueprint for one specific feature**, not the whole system.
-
-| Signal | Real human phrasing |
-|---|---|
-| Feature blueprint | "how do I implement X", "give me a blueprint for X", "design the X feature", "how should X work technically" |
-
-### `migration-guide` — Framework / language version upgrade
-
-| Signal | Real human phrasing |
-|---|---|
-| Upgrade | "upgrade from X to Y", "migrate to", "Next.js 12 → 15", "Python 2 → 3", "React 18 → 19" |
-| Major version | "major version bump", "breaking deps", "node 18 → 22", "TS 4 → 5", "moving to the new version" |
-| Deprecation | "this is deprecated", "the old way is removed", "I need to update away from X" |
-
----
-
-## Priority 6 — Testing
-
-### `tdd-guide` — Tests, coverage, TDD
-
-| Signal | Real human phrasing |
-|---|---|
-| Write tests | "write a test", "unit test", "integration test", "test for this", "I need tests for" |
-| TDD | "TDD", "test driven", "red green refactor", "write the test first" |
-| Coverage | "coverage", "what's not tested", "coverage gap", "my coverage is low", "what am I missing in tests" |
-| No tests yet | "I haven't written any tests", "there are no tests", "need to add tests" |
-
-### `e2e-runner` — End-to-end Playwright tests
-
-| Signal | Real human phrasing |
-|---|---|
-| E2E | "Playwright", "end-to-end", "E2E", "browser test", "user flow test", "test the whole flow" |
-| Critical path | "checkout flow", "signup flow", "happy path test", "test the full user journey" |
-
-### `pr-test-analyzer` — Test coverage on PRs
-
-| Signal | Real human phrasing |
-|---|---|
-| PR | "review my PR tests", "did I cover everything", "PR test gaps", "is my PR well tested" |
-
----
-
-## Priority 7 — Code Hygiene
-
-### `refactor-cleaner` — Dead code, cleanup
-
-| Signal | Real human phrasing |
-|---|---|
-| Cleanup | "clean up", "dead code", "unused", "remove old", "tech debt", "time to clean this up" |
-| Duplicate | "duplicate", "DRY", "extract function", "this is copy-pasted everywhere" |
-| Cruft | "old stuff hanging around", "deprecated code we never removed", "vestigial code" |
-
-### `code-simplifier` — Reduce complexity
-
-| Signal | Real human phrasing |
-|---|---|
-| Simplify | "simpler", "too complex", "hard to read", "convoluted", "this is too complicated" |
-| Readability | "easier to follow", "more readable", "nobody can understand this", "even I can't read it" |
-
-### `type-design-analyzer` — Type system improvements
-
-| Signal | Real human phrasing |
-|---|---|
-| Type design | "better types", "stricter types", "type safety", "discriminated union", "generics" |
-| `any` cleanup | "remove any", "stricter typing", "narrow type", "too many anys", "weaken the types" |
-
----
-
-## Priority 8 — Documentation
-
-### `doc-updater` — Update docs/README
-
-| Signal | Real human phrasing |
-|---|---|
-| Docs | "update README", "document this", "add docs", "JSDoc", "docstring", "write the docs for" |
-| Onboarding | "explain how this works", "doc for new devs", "I need to document this for my team" |
-| Missing docs | "nobody knows how this works", "there are no docs", "write documentation" |
-
-### `docs-lookup` — Library / framework API question
-
-| Signal | Real human phrasing |
-|---|---|
-| Library API | "how do I use X library", "what's the signature", "API for X", "how does X work" |
-| Framework feature | "Next.js App Router", "React Suspense", "Django ORM", "Rails 7", "how does X handle Y" |
-| Usage question | "what's the syntax for", "what parameters does X take", "how do I call" |
-
-### `comment-analyzer` — Comment quality
-
-| Signal | Real human phrasing |
-|---|---|
-| Comments | "are my comments good", "comment hygiene", "explain why not what", "are these comments useful" |
-
----
-
-## Priority 9 — Specialized Workflows
-
-### `chief-of-staff` — Communication, email, multi-channel ops
-
-Trigger for non-code workflow tasks: drafting emails, scheduling, status updates, multi-channel coordination.
-
-| Signal | Examples |
-|---|---|
-| Communication | "draft an email", "write a message", "reply to this", "slack message" |
-| Status | "status update", "what do I tell the team", "write the incident update" |
-
-### `seo-specialist` — SEO / metadata / search
-
-| Signal | Examples |
-|---|---|
-| SEO | "SEO", "meta tags", "schema.org", "structured data", "Core Web Vitals", "sitemap", "my page doesn't rank" |
-
-### `opensource-forker` + `opensource-sanitizer` + `opensource-packager` — Open-sourcing a project
-
-| Signal | Examples |
-|---|---|
-| OSS prep | "open source this", "make this public", "strip secrets", "remove credentials", "release as open source" |
-
-Use the chain: `forker` → `sanitizer` → `packager`.
-
-### `/dashboard` — Local observability dashboard
-
-| Signal | Examples |
-|---|---|
-| Open dashboard | "open the dashboard", "show me the dashboard", "start the dashboard" |
-| Inspect memory | "show me what's in my memory", "browse my sessions", "see my memory store" |
-| Token / evolve | "what's my token budget", "show evolve proposals", "evolve tab" |
-
-### `/swarm` — Parallel agent swarm in git worktrees
-
-| Signal | Examples |
-|---|---|
-| Swarm request | "use the swarm", "run a swarm", "swarm this", "parallel swarm" |
-| Parallel worktrees | "run agents in parallel", "spawn multiple agents", "maximum throughput" |
-
-### `/route-model` — Cost-aware model selection
-
-| Signal | Examples |
-|---|---|
-| Model choice | "which model should I use", "haiku or sonnet?", "should I use opus here" |
-| Cost concern | "cheapest model for this", "save tokens", "cost-aware routing" |
-
-### `/memory-evolve` — Self-evolving memory proposals
-
-| Signal | Examples |
-|---|---|
-| Evolve proposals | "show evolve proposals", "what has ECC learned", "memory evolution" |
-| Routing additions | "suggest routing improvements", "what routing misses happened" |
-
-### `/replay` — Session replay
-
-| Signal | Examples |
-|---|---|
-| Replay request | "replay a session", "re-run that session", "session replay" |
-| Bundle | "session bundle", "export and replay", "portable session" |
-
----
-
-## Priority 10 — Multi-Agent Patterns
-
-### Sequential chains (suggest the next agent)
-
-| Just used | Suggest next |
-|---|---|
-| `pair-programmer` (approach agreed) | `code-reviewer` after implementation |
-| `debug-detective` (root cause found) | `tdd-guide` to add a regression test |
-| `code-reviewer` (issues found) | `refactor-cleaner` to fix |
-| `security-reviewer` (vuln found) | `tdd-guide` to add a security test |
-| `architect` (design done) | `code-architect` for the first feature |
-| `performance-optimizer` (bottleneck found) | `tdd-guide` for a perf regression test |
-| `incident-commander` (incident resolved) | `debug-detective` for deeper root cause, then postmortem |
-
----
-
-### `image-architect` — Image and visual generation
-
-Trigger on ANY request for a visual, image, picture, graphic, or design asset.
-
-| Signal | Examples |
-|---|---|
-| Explicit request | "generate an image", "create a picture", "make a visual", "I need graphics" |
-| Hero / social | "hero image", "banner", "OG image", "social preview", "GitHub card", "Twitter card" |
-| Implicit visual need | "my site looks plain", "it looks empty", "needs something visual", "looks boring" |
-| Design ask | "design a logo", "make it look good", "something for the homepage" |
-
-Platform routing: Antigravity → Gemini Imagen 3 · Codex → DALL-E 3 · Claude Code → fal.ai → SVG · others → SVG.
-
----
-
-### `/project-launch` — New project parallel founding team
-
-Trigger when starting something new of meaningful scope.
-
-| Signal | Examples |
-|---|---|
-| New build | "help me build X", "I want to build X", "I'm building X", "let's make X" |
-| Greenfield | "starting from scratch", "new project", "new app", "new service", "new API" |
-| Idea phase | "I have this idea", "I'm thinking of building", "I want to make a" |
-| New SaaS / startup | "I'm building a SaaS", "side project", "startup idea" |
-
-**Key rule:** Non-trivial new build → `/project-launch`. Adding one feature to existing codebase → `pair-programmer`.
-
-Route to `/project-launch` — fires `architect` + `pair-programmer` + `security-reviewer` + `tdd-guide` + `ux-reviewer` simultaneously.
-
----
-
-### `/team-review` — Existing project parallel full audit
-
-Trigger when review scope is broad (multiple files) or the user asks "is this ready?".
-
-| Signal | Examples |
-|---|---|
-| Broad review | "review my project", "check my codebase", "audit my code" |
-| Pre-release | "before I ship", "ready to release?", "pre-launch check", "about to go live" |
-| Readiness check | "is this ready?", "any issues before I ship?", "is this production-ready?" |
-
-**Key rule:** Single-file review → language-specific reviewer. Full project or multi-file → `/team-review`.
-
-Route to `/team-review` — fires `code-reviewer` + `security-reviewer` + `performance-optimizer` + `api-guardian` simultaneously.
-
----
-
-### `/security-audit` — Full parallel security sweep
-
-Trigger when security concern spans the whole project or multiple vectors (secrets + deps + API).
-
-| Signal | Examples |
-|---|---|
-| Full security concern | "is my app secure?", "security audit", "run a security check" |
-| Pre-launch security | "before I go public", "about to expose this API" |
-| Dependency worry | "are my deps safe?", "any vulnerable packages?", "CVEs in my dependencies?" |
-
-**Key rule:** Single file concern → `security-reviewer`. Multi-vector or project-wide → `/security-audit`.
-
-Route to `/security-audit` — fires `security-reviewer` + `dependency-doctor` + `api-guardian` simultaneously.
-
----
-
-### `/devil-mode` — Full adversarial parallel sweep
-
-Trigger when attacker mindset is needed across the whole codebase, especially pre-public or post-compromise.
-
-| Signal | Examples |
-|---|---|
-| Attacker mindset | "think like an attacker", "red-team my code", "find what's exploitable" |
-| Pre-public sweep | "going open-source — full audit", "scan for everything before I publish" |
-| Post-compromise | "we were hacked, audit everything", "after the incident, full review" |
-| Devil mode keyword | "devil mode", "no-mercy audit", "paranoid review" |
-
-**Key rule:** One or two security angles → `/security-audit`. Multi-vector attacker sweep → `/devil-mode`.
-
-Route to `/devil-mode` — fires `prompt-injection-hunter` + `supply-chain-auditor` + `secret-hunter` + `backdoor-hunter`. Use `--all` for all 8, `--pre-public` for OSS sweep, `--pre-launch` for launch sweep.
-
----
-
-### `/pre-release` — Ship-readiness parallel check
-
-Trigger when cutting a release, tagging a version, or deploying to production.
-
-| Signal | Examples |
-|---|---|
-| About to release | "ready to cut v2?", "cutting a release", "tagging v1.0", "shipping today" |
-| Release checklist | "release checklist", "anything missing before v2?" |
-| Final sanity check | "last check before I push to prod", "one more look before shipping" |
-
-**Key rule:** General readiness → `/team-review`. Explicit release or version cut → `/pre-release`.
-
-Route to `/pre-release` — fires `release-captain` + `security-reviewer` + `code-reviewer` simultaneously.
-
----
-
-### `/debug-blitz` — Triple-agent parallel debug for stubborn bugs
-
-Trigger when a bug has resisted normal debugging or spans multiple layers (code + env + hidden failure).
-
-| Signal | Examples |
-|---|---|
-| Persistent bug | "been fighting this for hours", "nothing is working", "tried everything" |
-| Multi-layer problem | "works locally not in CI", "works on my machine", "only fails in prod" |
-| Frustration signal | "I give up", "losing my mind", "been stuck for X hours/days" |
-
-**Key rule:** First-time bug → `debug-detective`. Persistent, multi-layer, or frustration-coded → `/debug-blitz`.
-
-Route to `/debug-blitz` — fires `debug-detective` + `silent-failure-hunter` + `env-debugger` simultaneously.
-
----
-
-### `/refactor-sprint` — Full parallel refactor + type + test pass
-
-Trigger when cleanup spans a module or file, especially when multiple concerns (readability + types + tests) come together.
-
-| Signal | Examples |
-|---|---|
-| Big cleanup | "refactor this module", "this code is a mess", "technical debt cleanup" |
-| Multi-concern quality | "improve quality, types, and add tests", "full cleanup", "legacy code overhaul" |
-
-**Key rule:** Single function cleanup → `refactor-cleaner`. Module-wide with types and tests → `/refactor-sprint`.
-
-Route to `/refactor-sprint` — fires `refactor-cleaner` + `code-simplifier` + `type-design-analyzer` + `tdd-guide` simultaneously.
-
----
-
-### `/onboard` — New-joiner parallel codebase onboarding
-
-Trigger when someone is new to a codebase and needs to understand it.
-
-| Signal | Examples |
-|---|---|
-| New to codebase | "I just joined the team", "inherited this codebase", "taking over this repo" |
-| Explore unfamiliar code | "never seen this code before", "where do I even start?" |
-| Ramp-up request | "help me get up to speed", "ramp me up on this" |
-
-**Key rule:** Single "explain this function" → `code-explorer`. Full codebase orientation → `/onboard`.
-
-Route to `/onboard` — fires `code-explorer` + `architect` + `doc-updater` simultaneously.
-
----
-
-### `/lessons` — Project lessons management
-
-Trigger when the user references project rules, preferences, corrections, or past instructions.
-
-| Signal | Examples |
-|---|---|
-| Load rules | "load my lessons", "what rules do we have", "show me the rules" |
-| Remember something | "remember that we use pnpm", "note that we do X", "don't forget Y" |
-| Habit encode | "from now on always X", "make it a rule that", "I want you to always" |
-
-Proactively offer to load if `tasks/lessons.md` exists in the project root.
-
----
-
-### Parallel suggestions
-
-When the user's request spans multiple concerns, name the parallel agents:
-
-- "I'm building a payment endpoint" → `pair-programmer` + `security-reviewer` + `api-guardian`
-- "Refactoring the auth module" → `refactor-cleaner` + `security-reviewer` + `tdd-guide`
-- "help me build a SaaS app" → `/project-launch` (5 agents)
-- "review my project before I deploy" → `/team-review` (4 agents)
-- "my site needs visuals" → `image-architect`
-
----
-
-## Counter-Patterns — Do NOT Route
-
-**Skip routing if:**
-
-- The user **already explicitly invoked** an agent (`use <agent>`, `@agent`, `invoke <agent>`) — that takes precedence
-- The message is **a one-liner factual question** — answer directly
-- The message is **purely conversational** ("hi", "thanks", "ok") — respond normally
-- The user says **"just answer me directly"** or **"don't route"** — respect it
-- The request is **trivially simple** (rename a variable, change one line)
-
----
-
-## Example Routing Decisions
-
-| User says | Route to | Why |
-|---|---|---|
-| "I'm getting a TypeError on line 42" | `debug-detective` | Specific error |
-| "something is wrong but there's no error" | `silent-failure-hunter` | Silent failure pattern |
-| "Should I use React Context or Zustand here?" | `pair-programmer` | Pre-implementation approach |
-| "Review my login component" | `typescript-reviewer` | Single file + likely TS |
-| "I have no idea where to start" | `kodelyth-advisor` | Lost / overwhelmed |
-| "How do I make this faster?" | `performance-optimizer` | Direct perf question |
-| "Is my JWT signing secure?" | `security-reviewer` | Auth + security keyword |
-| "build failed on Vercel" | `build-error-resolver` | Build failure |
-| "Migrate from Pages Router to App Router" | `migration-guide` | Framework migration |
-| "Add accessibility to this form" | `ux-reviewer` | a11y |
-| "production is down, getting 500s" | `incident-commander` | Active production incident |
-| "will this hold under 10k users?" | `load-tester` | Capacity question |
-| "generate a hero image" | `image-architect` | Explicit image request |
-| "my site looks plain" | `image-architect` | Implicit visual need |
-| "help me build a todo app" | `/project-launch` | New build |
-| "I have this idea for a SaaS dashboard" | `/project-launch` | Greenfield project |
-| "review my code before I deploy" | `/team-review` | Pre-release, full scope |
-| "is my app secure? check everything" | `/security-audit` | Full security sweep |
-| "cutting v2 today, last check" | `/pre-release` | Release + confidence combo |
-| "been stuck on this bug for 2 days" | `/debug-blitz` | Persistent frustration |
-| "works locally not in prod" | `/debug-blitz` | Multi-layer problem |
-| "clean up this whole module, add types and tests" | `/refactor-sprint` | Multi-concern quality |
-| "I just joined this team, where do I start?" | `/onboard` | New-joiner onboarding |
-| "think like an attacker, sweep my code" | `/devil-mode` | Attacker mindset trigger |
-| "going open-source, full audit before publish" | `/devil-mode --pre-public` | Pre-public sweep |
-| "we were hacked, audit everything" | `/devil-mode --all` | Post-compromise full sweep |
-| "is my AI feature jailbreak-safe?" | `prompt-injection-hunter` | AI-feature attack surface |
-| "did I leak secrets in git history?" | `secret-hunter` | Credential leak hunt |
-| "remember we always use pnpm here" | `/lessons` | Preference to encode |
-| [user pastes code with no text] | `code-reviewer` | Implicit review |
-| [user pastes stack trace with no text] | `debug-detective` | Implicit debug |
-
----
-
-## Output Format When Routing
-
-**Default (verbose)**:
-```
-→ Routing to debug-detective (stack trace + frustration signals match the bug-tracking pattern)
-
-[response in debug-detective's style — methodical, hypothesis-driven, asks for repro]
-
-Tip: next time you can type "use debug-detective" to invoke me directly.
-```
-
-**Terse mode active (see Dimension 6)**:
-```
-→ debug-detective
-
-[response — no tip line, no explanation of why]
-```
-
-**Medium-confidence single match** (add a "wrong?" tail):
-```
-→ Routing to code-reviewer (single file paste, no other context — treating as review)
-
-[response as code-reviewer]
-
-Not what you wanted? Say "use debug-detective" or "just answer" to switch.
-```
-
-**Low-confidence, two candidates** (ask, do not pick):
-```
-Two agents fit this: debug-detective (bug + error) or silent-failure-hunter (no error thrown).
-Which one? Or should I answer directly?
-```
-
-**Sticky routing continuation** (already in an agent, no announcement):
-```
-[response continues in the same agent's voice — no re-announcement]
-```
-
-**Compound intent → parallel command**:
-```
-→ Firing /debug-blitz — 3 agents in parallel (debug-detective + silent-failure-hunter + env-debugger)
-
-[synthesised result]
-```
-
-This keeps the user **informed**, **never surprised**, and **learning the toolkit** with every interaction — while still respecting terse mode and session context.