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 instructions/ai-creed/ai-cortex/agents-mdgit clone --depth 1 https://github.com/ai-creed/ai-cortexWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/ai-creed/ai-cortex/agents-md)<a href="https://agentmods.dev/instructions/ai-creed/ai-cortex/agents-md"><img src="https://agentmods.dev/badge/instructions/ai-creed/ai-cortex/agents-md.svg" alt="Measured on agentmods" height="20"></a>What 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 | $0.01122 | $0.01122 |
| Opus 5 | $0.00561 | $0.00561 |
| Sonnet 5 | $0.00224 | $0.00224 |
| Haiku 4.5 | $0.00112 | $0.00112 |
Grade B, and why
ai-cortex AGENTS.md scanned grade B with 1 finding 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 4d ago.
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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
- Hook installation under `~/.claude/settings.json` and `~/.codex/config.toml` is the explicit exception — those are user-owned config (not repo-owned), and the user opts in via `ai-cortex history install-hooks`. How it starts
The opening of the file, as written. The whole thing — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS
Project Summary
ai-cortexis a local-first intelligence layer for AI agents — fast project rehydration, file discovery, call-graph awareness, captured session history, and a memory layer that records project rules and surfaces them at edit time.- Architecture is three-layer: structural (rehydrate / suggest / blast-radius), continuity (history + memory), and integration (MCP + briefing + adoption tooling).
- All persistent state lives under
~/.cache/ai-cortex/. ai-cortex MUST NOT write to user repository files.
Source Of Truth
- Tracked architecture and product decisions live in
docs/superpowers/specs/. - Local execution plans live in
docs/superpowers/plans/. docs/superpowers/plans/is git-ignored and local-only. Do not commit or re-track files from that directory.- User-facing docs:
README.mdplus the curated hierarchy underdocs/(getting-started,concepts,guides,reference,architecture, androadmap).
Repo Layout
src/cli.ts: CLI entrypoint.src/version.ts: canonical version source. Kept in lockstep withpackage.jsonbyscripts/release.sh.src/lib/: core library (indexer, retrievers, briefing, update-notifier, migration-notifier).src/lib/history/: session-history capture + hook installation.src/lib/memory/: memory layer (lifecycle, recall, scope match, surface hook, briefing digests).src/lib/stats/: telemetry sinks + readers used by the stats TUI and CLI report.src/mcp/server.ts: MCP server with the agent-facing tool surface.tests/unit/,tests/integration/: vitest test suites.scripts/release.sh,scripts/lib/release-headline.ts: release tooling.
Memory Layer Self-Use
- This project uses its own memory layer. Project-scoped memories surface during edits via the
PreToolUsehook and the rehydration briefing. - Do not put scars, gotchas, or recurring lessons in this file. Record them as memories via
record_memoryso they surface only when relevant (and don't bloat every agent's context). - Cardinal pattern:
recall_memoryis browse-only and does not signal usage;get_memory(id)is the consult signal that drives cleanup eligibility.
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.
- 4d ago First seen · 70 lines · 1,122 tokens per session scan B 3a655ab4c020
ai-cortex AGENTS.md is an instructions file published in the GitHub repository ai-creed/ai-cortex (7 stars, last pushed 1mo ago), licensed MIT. It adds 1,122 tokens to every session, about $0.0056 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
rag-rat AGENTS.md
Instructions for cq27-dev/rag-rat, covering agents.md, prefer the rag-rat mcp for code browsing, record durable learnings as rag-rat memories, repo orientation and worktree correctness.
prism CLAUDE.md
Instructions for provasign/prism, covering before you commit or release — run the regression suite and prism — context delivery.
vespertide AGENTS.md
AGENTS.md instructions for dev-five-git/vespertide, covering vespertide knowledge base, structure, where to look, data flow and conventions.
aeon CLAUDE.md
Instructions for aeonfun/aeon, covering aeon, how aeon works, strategy, voice and soul file hierarchy (read in this order).
mcp-structured-memory CLAUDE.md
Claude Code instructions for nmeierpolys/mcp-structured-memory, a project described as: Structured Memory MCP Server.
inkwell-memory CLAUDE.md
Instructions for veronchenko/inkwell-memory, covering claude.md — inkwellmemory, layout, multi-tenant mode (inkwellmultitenant=1), conventions and testing.