---
category: reference
tags: [claude-code, memory, process, backup]
last_updated: 2026-03-20
confidence: high
---

# Claude Code Project Memory Backup

Reorganized 2026-03-20. Memory is now tiered for progressive disclosure:

- **MEMORY.md** — always loaded. Behavioral rules inline, operational rule index with scopes.
- **Operational rule files** — loaded by scope when task matches. Tracked with `last_used` frontmatter.
- **archive.md** — cold operational rules, consulted on demand.
- **Wiki** — all informational/project knowledge. No local copies.

---

## MEMORY.md (Operational Briefing)

### Project
- robot.wtf — multi-tenant wiki platform (app server, auth, MCP, management API).
- Repos under `/Users/sderle/code/otterwiki/`. Parent dir is NOT a git repo.

### User
- Concise, no sycophancy, no fake enthusiasm.
- Bias toward minimal solutions. Don't over-engineer.
- Uses plan mode for non-trivial changes.

### Behavioral rules
- Opus orchestrates, Sonnet implements. Always.
- All subagents run in background, no exceptions.
- Worktrees must target a specific repo (e.g. `robot.wtf/`), never the parent directory.
- TDD by default. Discuss before skipping.
- Rule of Two: dispatch a second agent to independently verify any subagent finding before accepting it.
- Plan review before implementation. Insert explicit Loop Check tasks in the pipeline where the orchestrator decides whether to revise or proceed.
- Verify branch state with git log before any merge.
- Dispatch verification agent before merging subagent work.
- Fix known issues now. Don't carry debt.
- Don't ask questions you can answer by reading code, wiki, or docs.
- Question the premise before optimizing. Ask "should this exist?" before "how to make it faster?"
- Never ignore pre-existing test failures. Fix or flag.
- On session start, scan operational rules below. Load any that match the task.
- Follow Design/Implementation_Workflow in wiki for full pipeline.

### Memory management
- Behavioral rules: inline in MEMORY.md, no separate files.
- Operational rules: individual files with scope and last_used in frontmatter.
- Informational/project knowledge: write to dev wiki, not local files.
- Cold operational rules: move index entry to archive.md.
- Session end: update last_used on any operational rule that was applied.
- Review when: last_used > 2 weeks OR MEMORY.md exceeds 150 lines.

### Wiki knowledge base (query on demand via dev-wiki MCP)
- Design/ — feature specs, architecture decisions
- Dev/ — implementation summaries, benchmarks, retrospectives
- Tasks/ — task breakdowns (To-Do page is authoritative tracker)
- Security/ — OWASP audit
- Dev/Claude_Code_Memory_Backup — canonical memory snapshot
- Use semantic_search when the page location isn't obvious from the index.

### Operational rules (load by scope)
- repos-and-tools.md — scope: git
- vps-access.md — scope: deploy
- feedback_reproducible_vps.md — scope: deploy
- feedback_otterwiki_config_override.md — scope: db
- reference_db_locations.md — scope: db
- mcp-servers.md — scope: mcp
- mcp-auth.md — scope: mcp
- feedback_preflight_checklist.md — scope: workflow
- feedback_explicit_loop_checks.md — scope: workflow

---

## Operational Rule Files (full content)

### repos-and-tools.md (scope: git)

All repos live under `/Users/sderle/code/otterwiki/`. The parent directory is NOT a git repo.

**F/OSS Repos:**
- `otterwiki/` — Fork of redimp/otterwiki (wiki engine). Upstream: `github.com:redimp/otterwiki`. Our remote: `github.com:schuyler/otterwiki` (SSH). Deploy branch: `wikibot-io`. Feature branches: `feat/*` off `wikibot-io`.
- `otterwiki-api/` — REST API plugin. Remote: `github.com:schuyler/otterwiki-api`. Branch: `main`.
- `otterwiki-semantic-search/` — FAISS + ONNX MiniLM semantic search plugin. Remote: `github.com:schuyler/otterwiki-semantic-search`. Branch: `main`.
- `otterwiki-mcp/` — MCP server plugin (12 tools). Remote: `github.com:schuyler/otterwiki-mcp`. Branch: `main`.

**Private Repos:**
- `robot.wtf/` — Platform app server, auth, MCP sidecar, API, Ansible deploy. Remote: `github.com:schuyler/robot.wtf`.
- `wikibot-io/` — AWS SaaS predecessor (Pulumi IaC, Lambda packaging). Remote: `github.com:schuyler/wikibot-io`. AWS profile: `wikibot` (us-east-1).
- `deploy/` — Legacy 3GW deploy config. No git remote.
- `dev-wiki/` — Local dev wiki git repo.

**Deploy flow (robot.wtf):** `ansible-playbook ansible/deploy.yml -i ansible/inventory.yml` from robot.wtf repo.

### vps-access.md (scope: deploy)

**wiki-1 (app server):** `ssh sderle@192.168.77.107`
- Services: otterwiki (8000), MCP sidecar (8001), API (8002), auth (8003)
- Data: `/srv/data/robot.db`, `/srv/data/wikis/`, `/srv/data/robot.env`
- Ansible: `/Users/sderle/code/otterwiki/robot.wtf/ansible/`

**proxy-1 (Caddy):** `ssh sderle@robot.wtf`
- Read-only only. Managed by separate Ansible repo.

**3GW (local network):** LXC container at `3gw.lan`, zmx session `3gw`.

### feedback_reproducible_vps.md (scope: deploy)

**NEVER run SSH commands that mutate the VPS.** Allowed: read-only diagnostics (`journalctl`, `systemctl status`, `curl`, `cat`, `ls`, `sqlite3 SELECT`). NOT allowed: `pip install`, `sed -i`, `systemctl restart`, `rm`, `mkdir`, or any state-changing command.

Correct workflow: Edit Ansible role → commit → `ansible-playbook deploy.yml` → verify via read-only SSH.

Corrected five times. The 45-second deploy cycle is the cost of reproducibility and it is always worth paying. Subagents must also follow this rule.

### feedback_otterwiki_config_override.md (scope: db)

Otterwiki's `update_app_config()` loads preferences from the active DB and overwrites Flask config. Load order: defaults → settings.cfg → env vars → DB (DB wins). The startup default DB creates empty-string rows that overwrite settings.cfg.

Always seed config values in per-wiki preferences table via `_init_wiki_db()`. Do NOT rely on settings.cfg for user-modifiable config.

### reference_db_locations.md (scope: db)

**Platform DB** (`/srv/data/robot.db`): Schema in `ansible/roles/database/files/schema.sql`. Tables: `users`, `wikis`, `oauth_*`, `mcp_tokens`. Code: `app/db.py`, `app/models/`.

**Per-wiki DBs** (`/srv/data/wikis/{slug}/wiki.db`): Otterwiki's own SQLAlchemy schema. Tables: `preferences`, `user`, drafts. Code: `app/resolver.py` (`_init_wiki_db`). Initialized on first access (idempotent).

These are NOT the same schema. Code that writes to one does not affect the other.

### mcp-servers.md (scope: mcp)

Three MCP connections. Do NOT confuse them.

- **dev-wiki** (`mcp__dev-wiki__*`) = dev.robot.wtf — platform dev instance
- **thirdgulfwar** (`mcp__thirdgulfwar__*`) = 3gw.robot.wtf — research wiki (LXC/Caddy)
- **claude_ai_Third_Gulf_War_research_wiki** — same as thirdgulfwar, via Claude.ai connector

### mcp-auth.md (scope: mcp)

`MultiAuth` composes `InMemoryOAuthProvider` (OAuth 2.1 for Claude.ai) + `StaticTokenVerifier` (bearer token for Claude Code). OAuth is the transport layer — Caddy basic auth on `/authorize*` is the actual gate. Tokens are in-memory; server restart forces re-auth.

Key gotchas: Claude.ai does NOT support bearer tokens. DCR disabled by default. Claude.ai URL must include `/mcp` path. `StaticTokenVerifier` (not `DebugTokenVerifier`) for bearer fallback.

### feedback_preflight_checklist.md (scope: workflow)

Before dispatching ANY worktree-isolated agent:
1. Verify CWD is inside the target git repo
2. Confirm plan-review loop is complete (no FAIL or CONDITIONS outstanding)
3. Confirm task list dependencies are wired up

### feedback_explicit_loop_checks.md (scope: workflow)

Pipeline template:
```
- [ ] Plan
- [ ] Plan Review (blocked by Plan)
- [ ] Plan Loop Check (blocked by Plan Review) ← orchestrator evaluates
- [ ] Implement (blocked by Plan Loop Check)
- [ ] Chico review (blocked by Implement)
- [ ] Zeppo review (blocked by Implement)
- [ ] Review Loop Check (blocked by Chico + Zeppo) ← orchestrator evaluates
- [ ] Verify (blocked by Review Loop Check)
```

No standing "Fix" task. Fix tasks created only if Loop Check determines they're needed. Loop Check tasks completed by the orchestrator, not a subagent.