Feature/slash commands updates 11 2025

.cursor/plans/tres-2279f097.plan.md

@@ -0,0 +1,59 @@
+<!-- 2279f097-4e2b-4f96-80cc-5e90782fecd3 0e269a2f-0e5b-4bde-873d-12274991a4cd -->
+# Claude Code Tresor Enhancement Roadmap
+
+## Foundations & Memory Bank
+
+- Stand up mandatory memory bank (`projectbrief.md`, `productContext.md`, `activeContext.md`, etc.) to persist architecture, tech stack, and progress context.
+- Capture current taxonomy of skills, agents, commands, and scripts so all follow-on decisions share the same source of truth.
+
+## Metadata & Packaging Alignment
+
+- Reconcile installer expectations with actual asset formats by either generating `agent.json`/`command.json` files or teaching `scripts/install.sh` to consume the Markdown frontmatter so agents deploy correctly.
+```130:147:scripts/install.sh
+        # Copy all agent directories
+        find "$agents_src" -mindepth 1 -maxdepth 1 -type d | while read -r agent_dir; do
+            local agent_name=$(basename "$agent_dir")
+
+            # Skip README-only directories
+            if [ -f "$agent_dir/agent.json" ]; then
+                log "Installing agent: $agent_name"
+                cp -r "$agent_dir" "$agents_dest/$agent_name"
+            fi
+        done
+```
+
+- Audit documentation vs. real inventory so directories like `commands/README.md` stop advertising non-existent commands (e.g. `/refactor`, `/deploy`).
+```29:47:commands/README.md
+- **`/refactor`** - Automated code refactoring with best practices
+- **`/optimize`** - Performance analysis and optimization suggestions
+...
+- **`/deploy`** - Deployment automation and verification
+```
+
+- Generate machine-readable indexes (YAML/JSON) for skills, agents, and subagents to power search, validation, and future tooling.
+
+## Ecosystem & Automation Expansion
+
+- Design a metadata-driven CLI (e.g. `tresor`) that can list, preview, install, and update skills/agents/commands/subagents from the repo and the `sources/` library.
+- Introduce reusable command/agent templates so contributors can scaffold new utilities with validation hooks.
+- Build orchestration blueprints showing how autonomous skills feed subagents and commands for key workflows (code review, release, incident response).
+
+## Documentation & Adoption Experience
+
+- Produce a role-based navigation hub that maps typical Claude Code personas (solo dev, team lead, agency) to specific utilities and workflows.
+- Collapse duplicated docs (e.g. multiple READMEs per feature) into concise playbooks with deep links to authoritative references.
+- Add outcome-focused quick-starts and video-ready scripts that demonstrate end-to-end flows using bundled assets.
+
+## Quality Assurance & Release Management
+
+- Add automated tests/linters for installer scripts and metadata consistency; validate every PR via CI.
+- Capture usage analytics & feedback loops (changelogs, release notes, telemetry opt-in) to guide prioritization.
+- Formalize versioning and release cadence with signed tags, update scripts, and migration notes for breaking changes.
+
+## Implementation Todos
+
+- baseline: Establish memory bank & repository audit
+- packaging: Align installer, metadata, and docs with actual assets
+- automation: Build CLI tooling and workflow orchestration patterns
+- docs: Redesign documentation & onboarding experience
+- quality: Add automated validation, testing, and release process

.gitignore

@@ -127,3 +127,11 @@ scripts/publish-gist.sh
 
 documentation/archive/*
 # docs/ - NOW TRACKED: Technical documentation for agent system (v2.5.0+)
+__pycache__/
+
+# TÂCHES workflow data files (per-project data, not commands - v2.6.5+)
+# These are gitignored in USER projects, not in this repository
+# Pattern: Ignore in root and subdirectories, but not in commands/
+/*/.prompts/
+/TO-DOS.md
+/whats-next.md