tribunal-kit 3.0.0 → 4.0.0

package/.agent/workflows/debug.md

@@ -1,189 +1,165 @@
----
-description: Debugging command. Activates DEBUG mode for systematic problem investigation using the 4-phase methodology (Collect → Hypothesize → Test → Fix). No fix is suggested until the root cause is confirmed and tested. No random changes. No guessing.
----
-
-# /debug — Root Cause Investigation
-
-$ARGUMENTS
-
----
-
-## The Investigation Contract
-
-> "A fix without a root cause is a patch on a symptom. It will fail again."
-
-The `debugger` agent follows this sequence **without skipping steps**:
-
-1. Collect all evidence first
-2. Generate ranked hypotheses 
-3. Test exactly one hypothesis at a time
-4. Confirm root cause with evidence
-5. Apply the minimum targeted fix
-6. Add a regression test to prevent recurrence
-
----
-
-## When to Use /debug
-
-| Use `/debug` when... | Use something else when... |
-|:---|:---|
-| There's a specific error or unexpected behavior | Code needs to be written → `/generate` |
-| You have a stack trace or error message | Code quality needs improving → `/refactor` |
-| Production is broken right now | Missing test coverage → `/test` |
-| A bug reappears after being "fixed" | Full health check needed → `/audit` |
-
----
-
-## Step 1 — Evidence Collection (Collect All Before Hypothesizing)
-
-```
-□ Exact error text — full stack trace, not a paraphrase
-□ Minimum reproduction steps — fewest actions that trigger the bug
-□ Last known-good state — commit hash, date, or config snapshot
-□ Recent changes — code, deps, env vars, infra, config changes
-□ Environment — local/staging/prod, OS, Node version, browser, runtime
-□ Frequency — always / intermittent / only under load / production only
-```
-
-> ⚠️ If the error is intermittent: collect timing patterns before hypothesizing.
-
-### Priority Investigation Order (Most Likely Root Cause First)
-
-```
-1. Recent deployments     — 90% of outages are caused by recent changes
-2. Environment variables  — rotated/missing secrets are common silent failures
-3. Dependency updates     — a package update can break APIs without errors
-4. Infrastructure         — firewall, Security Groups, DB connection limits
-5. Application code       — last to check, easiest to blame prematurely
-```
-
----
-
-## Step 2 — Hypothesis Formation
-
-Map all possible causes with explicit likelihood labels:
-
-```
-ROOT CAUSE CANDIDATES
-━━━━━━━━━━━━━━━━━━━━━
-H1 [High]   — [cause] — Evidence: [what points to this]
-H2 [Medium] — [cause] — Evidence: [what is consistent with this]
-H3 [Low]    — [cause] — Evidence: [theoretically possible]
-```
-
-**Never state a hypothesis as confirmed fact until Step 3 proves it.**
-
----
-
-## Step 3 — Single-Hypothesis Testing
-
-Test **one at a time**. Never test two simultaneously — results become ambiguous.
-
-```
-H1 tested: [how it was investigated]
-Result:    ✅ Confirmed | ❌ Ruled out — [specific evidence against]
-
-H2 tested: [how it was investigated]
-Result:    ✅ Confirmed | ❌ Ruled out — [reason]
-```
-
-Stop when the first hypothesis is confirmed. Do not continue investigating eliminated causes.
-
----
-
-## Step 4 — Root Cause Statement
-
-The root cause is one sentence: **WHY this happened**, not WHAT happened.
-
-```
-✅ "JWT verification was skipped when Authorization header used lowercase 'bearer'
-    because the header check was case-sensitive"
-
-❌ "The API returned 401" — This is the symptom, not the root cause
-```
-
----
-
-## Step 5 — Fix + Regression Prevention
-
-```
-Targeted fix:    The minimum change that eliminates the root cause
-Regression test: A specific test that will catch this exact failure if it returns
-Similar patterns: Other locations in the codebase to audit for the same issue
-Debug cleanup:   All console.log/temporary changes removed from proposed fix
-```
-
----
-
-## Debug Report Format
-
-```
-━━━ Debug Report ━━━━━━━━━━━━━━━━━━━━━━━━━━
-
-Symptom:      [what the user observes]
-Error:        [exact message / stack trace]
-Reproduced:   Yes | No | Sometimes — [conditions]
-Environment:  [Node v22, Next.js 15, PostgreSQL 16]
-Last working: [commit hash / date]
-
-━━━ Evidence ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
-- [specific observation 1]
-- [specific observation 2]
-
-━━━ Hypotheses ━━━━━━━━━━━━━━━━━━━━━━━━━━
-H1 [High]   — [cause and reasoning]
-H2 [Medium] — [cause and reasoning]
-
-━━━ Investigation ━━━━━━━━━━━━━━━━━━━━━━━
-H1: [what was tested] → ✅ Confirmed root cause
-H2: [what was tested] → ❌ Ruled out — [reason]
-
-━━━ Root Cause ━━━━━━━━━━━━━━━━━━━━━━━━━
-[Single sentence WHY — not WHAT]
-
-━━━ Fix ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
-Before: [original code]
-After:  [corrected code]
-
-Regression test: [test preventing recurrence]
-Similar patterns: [other locations to check]
-```
-
----
-
-## Hallucination Guard
-
-```
-❌ Never propose a fix before root cause is confirmed with evidence
-❌ Never test two hypotheses simultaneously
-❌ Never propose a "rewrite" as a debug session
-❌ Never leave console.log in the proposed fix
-❌ Never assume the error message accurately describes the actual cause
-❌ Never use real API methods without verifying they exist in this version
-```
-
----
-
-## Cross-Workflow Navigation
-
-| After /debug shows... | Go to |
-|:---|:---|
-| Root cause confirmed, fix ready | `/generate` to write fix through Tribunal |
-| Multiple files need changing | `/enhance` for impact-zone analysis |
-| Missing test allowed this bug in | `/test` to add regression coverage |
-| Performance was the root cause | `/tribunal-performance` |
-| Security vulnerability found | `/audit` to find other instances |
-
----
-
-## Usage Examples
-
-```
-/debug TypeError: Cannot read properties of undefined reading 'id'
-/debug API returns 500 only in production environment
-/debug useEffect fires on every render instead of only once
-/debug login works locally but fails in CI pipeline
-/debug memory usage grows unbounded after 24 hours
-/debug race condition in payment confirmation handler
-/debug Next.js 15 cookies() throws "Dynamic server error"
-```
+---
+description: Debugging command. Activates DEBUG mode for systematic problem investigation using the 4-phase methodology (Collect → Hypothesize → Test → Fix). No fix is suggested until the root cause is confirmed and tested. No random changes. No guessing.
+---
+
+# /debug — Root Cause Investigation
+
+$ARGUMENTS
+
+---
+
+## The Investigation Contract
+
+"A fix without a root cause is a patch on a symptom. It will fail again."
+
+The `debugger` agent follows this sequence **without skipping steps**:
+
+1. Collect all evidence first
+2. Generate ranked hypotheses 
+3. Test exactly one hypothesis at a time
+4. Confirm root cause with evidence
+5. Apply the minimum targeted fix
+6. Add a regression test to prevent recurrence
+
+---
+
+## When to Use /debug
+
+|Use `/debug` when...|Use something else when...|
+|:---|:---|
+|There's a specific error or unexpected behavior|Code needs to be written → `/generate`|
+|You have a stack trace or error message|Code quality needs improving → `/refactor`|
+|Production is broken right now|Missing test coverage → `/test`|
+|A bug reappears after being "fixed"|Full health check needed → `/audit`|
+
+---
+
+## Step 1 — Evidence Collection (Collect All Before Hypothesizing)
+
+```
+□ Exact error text — full stack trace, not a paraphrase
+□ Minimum reproduction steps — fewest actions that trigger the bug
+□ Last known-good state — commit hash, date, or config snapshot
+□ Recent changes — code, deps, env vars, infra, config changes
+□ Environment — local/staging/prod, OS, Node version, browser, runtime
+□ Frequency — always / intermittent / only under load / production only
+```
+
+> ⚠️ If the error is intermittent: collect timing patterns before hypothesizing.
+
+### Priority Investigation Order (Most Likely Root Cause First)
+
+```
+1. Recent deployments     — 90% of outages are caused by recent changes
+2. Environment variables  — rotated/missing secrets are common silent failures
+3. Dependency updates     — a package update can break APIs without errors
+4. Infrastructure         — firewall, Security Groups, DB connection limits
+5. Application code       — last to check, easiest to blame prematurely
+```
+
+---
+
+## Step 2 — Hypothesis Formation
+
+Map all possible causes with explicit likelihood labels:
+
+```
+ROOT CAUSE CANDIDATES
+━━━━━━━━━━━━━━━━━━━━━
+H1 [High]   — [cause] — Evidence: [what points to this]
+H2 [Medium] — [cause] — Evidence: [what is consistent with this]
+H3 [Low]    — [cause] — Evidence: [theoretically possible]
+```
+
+**Never state a hypothesis as confirmed fact until Step 3 proves it.**
+
+---
+
+## Step 3 — Single-Hypothesis Testing
+
+Test **one at a time**. Never test two simultaneously — results become ambiguous.
+
+```
+H1 tested: [how it was investigated]
+Result:    ✅ Confirmed | ❌ Ruled out — [specific evidence against]
+
+H2 tested: [how it was investigated]
+Result:    ✅ Confirmed | ❌ Ruled out — [reason]
+```
+
+Stop when the first hypothesis is confirmed. Do not continue investigating eliminated causes.
+
+---
+
+## Step 4 — Root Cause Statement
+
+The root cause is one sentence: **WHY this happened**, not WHAT happened.
+
+```
+✅ "JWT verification was skipped when Authorization header used lowercase 'bearer'
+    because the header check was case-sensitive"
+
+❌ "The API returned 401" — This is the symptom, not the root cause
+```
+
+---
+
+## Step 5 — Fix + Regression Prevention
+
+```
+Targeted fix:    The minimum change that eliminates the root cause
+Regression test: A specific test that will catch this exact failure if it returns
+Similar patterns: Other locations in the codebase to audit for the same issue
+Debug cleanup:   All console.log/temporary changes removed from proposed fix
+```
+
+---
+
+## Debug Report Format
+
+```
+━━━ Debug Report ━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+Symptom:      [what the user observes]
+Error:        [exact message / stack trace]
+Reproduced:   Yes | No | Sometimes — [conditions]
+Environment:  [Node v22, Next.js 15, PostgreSQL 16]
+Last working: [commit hash / date]
+
+━━━ Evidence ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+- [specific observation 1]
+- [specific observation 2]
+
+━━━ Hypotheses ━━━━━━━━━━━━━━━━━━━━━━━━━━
+H1 [High]   — [cause and reasoning]
+H2 [Medium] — [cause and reasoning]
+
+━━━ Investigation ━━━━━━━━━━━━━━━━━━━━━━━
+H1: [what was tested] → ✅ Confirmed root cause
+H2: [what was tested] → ❌ Ruled out — [reason]
+
+━━━ Root Cause ━━━━━━━━━━━━━━━━━━━━━━━━━
+[Single sentence WHY — not WHAT]
+
+━━━ Fix ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+Before: [original code]
+After:  [corrected code]
+
+Regression test: [test preventing recurrence]
+Similar patterns: [other locations to check]
+```
+
+---
+
+## Hallucination Guard
+
+```
+❌ Never propose a fix before root cause is confirmed with evidence
+❌ Never test two hypotheses simultaneously
+❌ Never propose a "rewrite" as a debug session
+❌ Never leave console.log in the proposed fix
+❌ Never assume the error message accurately describes the actual cause
+❌ Never use real API methods without verifying they exist in this version
+```
+
+---