fix(memory): enforce lean index format, remove ephemeral cache files

## CRITICAL: Index File Format

Index files (skills-*-index.md) MUST contain ONLY the table:
- No headers, no descriptions, no metadata
- Maximum token efficiency

Stripped all 30 index files to table-only format.

## Cache Strategy Update

Removed ephemeral cache files from git:
- github-open-prs-cache.md (deleted)
- github-open-issues-cache.md (deleted)

Reason: Cache files in git would cause merge conflicts and slow merge velocity.

Recommendation: Use session-local or cloudmcp caching instead.

## Agent Documentation

Added CRITICAL guidance to memory.md, skillbook.md, and shared templates
about index file format requirements.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

Author: rjmurillo-bot

.serena/memories/github-open-issues-cache.md

@@ -6,17 +6,37 @@
 - **Indexed Files**: 197 atomic skills
 - **Token Efficiency**: 82% savings with session caching
 
-## Proposed Cache-Aside Memory Domains
+## Cache Strategy (DECISION NEEDED)
 
-### Priority 1: High-Value Caches (Reduce GitHub API calls)
+### Problem
 
-| Domain | Cache Target | TTL | Invalidation Trigger |
-|--------|--------------|-----|----------------------|
-| `github-open-issues-cache` | Open issues list | 1 hour | Issue created/closed webhook |
-| `github-open-prs-cache` | Open PRs list | 30 min | PR opened/merged/closed |
-| `github-labels-milestones-cache` | Labels and milestones | 24 hours | Manual or on change |
+Ephemeral cache files (open PRs, open issues) in git would:
+1. Cause merge conflicts on every PR
+2. Slow merge velocity
+3. Require constant conflict resolution from main
 
-**Benefit**: Agents frequently query open issues/PRs. Caching saves 10-20 API calls per session.
+### Options
+
+| Option | Pros | Cons |
+|--------|------|------|
+| **Session-local cache** | No merge conflicts | No cross-session benefit |
+| **cloudmcp cache** | Cross-session, no conflicts | External dependency |
+| **Invalidate-on-write** | Fresh data guaranteed | Requires write hooks |
+| **No cache (query API)** | Always fresh | Higher API usage |
+
+### Recommendation
+
+Do NOT store ephemeral cache data in `.serena/memories/`. Use one of:
+1. Session-local variables (in-context)
+2. cloudmcp for cross-session caching
+3. Local temp files outside git tracking
+
+### Invalidation Strategy
+
+Prefer invalidate-on-write over TTL-on-read:
+- When closing PR/Issue: invalidate cache
+- When creating PR/Issue: invalidate cache
+- Avoids stale data from TTL-based reads
 
 ### Priority 2: Architecture Reference (Rarely Changes)
 

.serena/memories/github-open-prs-cache.md

@@ -41,15 +41,6 @@
 | cva refactoring variant consolidation template generate | skills-cva-refactoring |
 | phase3 agent skill session handoff template verification | skills-agent-workflow-phase3 |
 
-## GitHub Cache (Cache-Aside Pattern)
-
-| Cache Keywords | Memory File | TTL |
-|----------------|-------------|-----|
-| open pr list pull request ready review | github-open-prs-cache | 30 min |
-| open issue list bug enhancement priority | github-open-issues-cache | 1 hour |
-
-**Usage**: Read cache first. If stale (check timestamp), refresh from API and update memory.
-
 ## User Constraints (MUST READ)
 
 | Constraint | Memory |

.serena/memories/issue-307-memory-automation.md

@@ -1,11 +1,3 @@
-# Agent Workflow Domain Index
-
-**Purpose**: Route to atomic skills for multi-agent pipelines, scope discipline, and execution patterns.
-
-**Source**: Linting session, PR43 retrospective
-
-## Activation Vocabulary
-
 | Keywords | File |
 |----------|------|
 | pipeline analyst architect planner critic implementer qa full | agent-workflow-pipeline |
@@ -14,14 +6,3 @@
 | scope discipline discovery defer document expand mid-execution | agent-workflow-scope-discipline |
 | MVP ship working perfect gold-plating tech debt constraint | agent-workflow-mvp-shipping |
 | collaboration user addition scope gap signal learning heuristic | agent-workflow-collaboration |
-
-## Domain Statistics
-
-- **Skills**: 6 core patterns
-- **Anti-Patterns**: 3 documented
-- **Pipeline Stages**: 7
-
-## See Also
-
-- `skills-orchestration-index` - Parallel execution coordination
-- `skills-quality-index` - QA routing patterns

.serena/memories/memory-index.md

@@ -1,23 +1,5 @@
-# Analysis Domain Index
-
-**Purpose**: Route to atomic skills for capability gap analysis, comprehensive analysis standards, and root cause investigation.
-
-**Source**: `.agents/analysis/` patterns
-
-## Activation Vocabulary
-
 | Keywords | File |
 |----------|------|
 | gap analysis template severity root cause remediation affected | analysis-gap-template |
 | comprehensive analysis options trade-off recommendation specs | analysis-comprehensive-standard |
 | git blame root cause commit PR investigation history context | analysis-git-blame |
-
-## Domain Statistics
-
-- **Skills**: 3 core skills
-- **Validation**: Sessions 19-21 (100% implementation accuracy)
-
-## See Also
-
-- `skills-critique` - Conflict escalation
-- `skills-planning-index` - Parallel exploration

.serena/memories/skills-agent-workflow-index.md

@@ -1,24 +1,6 @@
-# Architecture Domain Index
-
-**Purpose**: Route to atomic skills for ADRs, tool allocation, model selection, and composite actions.
-
-**Source**: `.agents/architecture/` ADRs
-
-## Activation Vocabulary
-
 | Keywords | File |
 |----------|------|
 | tool allocation role-specific agent blanket security implementer analyst | architecture-tool-allocation |
 | model selection opus sonnet haiku tier complexity cost | architecture-model-selection |
 | composite action reusable workflow DRY extract parameterized | architecture-composite-action |
 | producer consumer prompt coordination handoff format output | architecture-producer-consumer |
-
-## Domain Statistics
-
-- **Skills**: 4 core skills
-- **Sources**: ADR-0002, ADR-0003, Session 03, Session 17
-
-## See Also
-
-- `skills-design-index` - Agent design principles
-- `skills-orchestration-index` - Agent coordination patterns

.serena/memories/skills-analysis-index.md

@@ -1,24 +1,5 @@
-# Bash Integration Domain Index
-
-**Purpose**: Route to atomic skills for bash script development, cross-language exit codes, and pre-commit hook patterns.
-
-**Source**: PR #52 retrospective
-
-## Activation Vocabulary
-
 | Keywords | File |
 |----------|------|
 | bash pattern discovery AUTOFIX pre-commit search grep section | bash-pattern-discovery |
 | exit code return PowerShell script scope cross-language contract | bash-exit-codes |
 | LASTEXITCODE pester integration test cross-language invoke | bash-exit-code-testing |
-
-## Domain Statistics
-
-- **Skills**: 5 core skills
-- **Key Pattern**: exit 1 (not return) in script scope
-
-## See Also
-
-- `skills-pester-testing` - Pester test patterns
-- `git-hook-patterns` - Git hook patterns
-- `skills-powershell-index` - PowerShell skills

.serena/memories/skills-architecture-index.md

@@ -1,11 +1,3 @@
-# CI Infrastructure Skills Index
-
-**Domain**: CI/CD pipelines, GitHub Actions, build automation
-**Skills**: 20
-**Updated**: 2025-12-23
-
-## Activation Vocabulary
-
 | Keywords | File |
 |----------|------|
 | pester test runner invoke artifacts output xml junit nunit CI local | ci-test-runner-artifacts |
@@ -24,9 +16,6 @@
 | integration test workflow dry-run module import validation pre-merge | ci-quality-gates |
 | research platform documentation limitation compatibility known issue | ci-deployment-validation |
 | label pre-validation create ensure exists gh label create deploy | ci-deployment-validation |
-
-## Coverage
-
 | File | Skills | Focus |
 |------|--------|-------|
 | ci-test-runner-artifacts | TestRunner-001, Artifacts-001 | Test execution patterns |
@@ -38,9 +27,3 @@
 | ci-ai-integration | Structured-Output-001, Comment-001 | AI automation |
 | ci-quality-gates | CI-001, Infra-003, Integration-001 | Validation gates |
 | ci-deployment-validation | Research-002, Infra-004 | Pre-deployment checks |
-
-## Related
-
-- Source: Sessions 03-08, PRs #47, #79, #202, #212
-- GitHub workflow patterns: `skills-github-workflow-patterns`
-- PowerShell CI: `powershell-cross-platform-ci`

.serena/memories/skills-bash-integration-index.md

@@ -1,11 +1,3 @@
-# CodeRabbit Domain Index
-
-**Purpose**: Route to atomic skills for CodeRabbit configuration, false positive patterns, and path-based filtering.
-
-**Consolidated Sources**: coderabbit-config-strategy, coderabbit-path-instructions, coderabbit-false-positives
-
-## Activation Vocabulary
-
 | Keywords | File |
 |----------|------|
 | config profile chill assertive path-filter noise reduction yaml | coderabbit-config-strategy |
@@ -14,8 +6,3 @@
 | infrastructure naming spaces security injection vulnerable expression | coderabbit-security-false-positives |
 | generated files edit warning analyst impact-analysis nested fence backtick | coderabbit-documentation-false-positives |
 | markdownlint MD004 MD012 MD013 duplicate linting yaml | coderabbit-markdownlint |
-
-## Domain Statistics
-
-- **Skills**: 6 atomic skills
-- **Validated**: 12 PRs