Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add rules/iamtouchskyer/memex/cursorrulesgit clone --depth 1 https://github.com/iamtouchskyer/memexWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.01610 | $0.01610 |
| Opus 5 | $0.00805 | $0.00805 |
| Sonnet 5 | $0.00322 | $0.00322 |
| Haiku 4.5 | $0.00161 | $0.00161 |
Grade A, and why
cursorrules scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 132 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Memex — Agent Working Instructions
Detailed architecture reference: docs/ARCHITECTURE.md
What Is This
Zettelkasten memory for AI agents. Markdown cards in ~/.memex/cards/ with [[wikilinks]]. npm package @touchskyer/memex. Distributed as CLI + MCP server + Claude Code plugin + VS Code extension + Pi extension.
Before You Touch Anything
npm test # must pass. vitest, 70% coverage threshold.
Build has a known TS error (node-llama-cpp optional dep). Ignore it — it compiles fine for distribution.
How The Code Is Layered
MCP tools (src/mcp/) → Commands (src/commands/) → Lib (src/lib/) → Filesystem (~/.memex/)
↑ client-facing ↑ business logic ↑ data access
The rule: MCP tools and CLI are thin wrappers. All logic lives in src/commands/. All data access in src/lib/store.ts. Don't put logic in the wrong layer.
Files You'll Actually Need
| When you're changing... | Read these first |
|---|---|
| Card read/write/slug behavior | src/lib/store.ts (164 lines — read all of it) |
| MCP tool behavior | src/mcp/operations.ts (high-level), src/mcp/server.ts (low-level) |
| Search (keyword or semantic) | src/commands/search.ts (most complex file, ~390 lines) |
| Sync / git operations | src/lib/sync.ts (GitAdapter, autoSync, autoFetch) |
| Embeddings | src/lib/embeddings.ts (3 providers + cache) |
| Card parsing / frontmatter | src/lib/parser.ts (47 lines — tiny but critical) |
| Organize / graph analysis | src/commands/organize.ts |
| Claude Code plugin hooks | hooks/hooks.json |
| VS Code extension | vscode-extension/src/extension.ts |
| Pi extension | pi-extension/index.ts |
Do / Don't
DO
- Keep commands pure: return
{output, exitCode}or{success, content/error}. Let CLI/MCP handle I/O. - Use
.jsextensions in all imports (Node16 ESM resolution). - Test both
nestedSlugs: false(default) andnestedSlugs: truewhen touchingstore.ts. - Run
npm testbefore declaring anything done. - Add new MCP tools in
operations.tsif they need hooks, inserver.tsif they don't. - Use
zodfor MCP tool input schemas (already imported in both MCP files).
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 132 lines · 1,610 tokens per session scan A 7851fffd5a91
cursorrules is a cursor rule published in the GitHub repository iamtouchskyer/memex (140 stars, last pushed 1mo ago), licensed MIT. It adds 1,610 tokens to every session, about $0.0080 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other cursor rules, from other repositories
cursor-tools-mastery
Cursor 3.7 runtime guide: choose the right tool, canvases, Design Mode, /worktree, /best-of-n, Await, and parallel execution where safe.
cursor-agent-orchestration
Cursor 3.7 orchestration guide: when to plan, when to delegate, nested subagents, multi-environment handoffs, /best-of-n, and Await for long-running branches.
fable5-reasoning
Fable 5 reasoning protocols: task interpretation, risk-first decomposition, approach selection, interleaved thinking, hypothesis ledgers, premortems, calibration, and the stuck-strategy ladder. Load for complex, ambiguous, or long-horizon tasks, for debugging strategy, or whenever progress stalls.
cursor-mcp-optimization
Cursor 3.7 MCP optimization: browser Design Mode, canvases, Figma, Cloudflare tools, MCP Apps structured content, and direct action patterns.
minimax-mcp-tools
MCP and web-tool guidance: current-doc retrieval, direct-tool preference, MCP Apps structured content, and version-aware external lookups.
reqforge-dev-build
Dev Builder — implementation execution. Build code phase by phase following DEV-PLAN, each Task goes through review → fix loop.