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 skills/vikasgrac/project-brain/recallnpx skills add vikasgrac/project-brain --skill recallgit clone --depth 1 https://github.com/vikasgrac/project-brainWrote 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/skills/vikasgrac/project-brain/recall)<a href="https://agentmods.dev/skills/vikasgrac/project-brain/recall"><img src="https://agentmods.dev/badge/skills/vikasgrac/project-brain/recall.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 | $0.00040 | $0.00540 |
| Opus 5 | $0.00020 | $0.00270 |
| Sonnet 5 | $0.00008 | $0.00108 |
| Haiku 4.5 | $0.00004 | $0.00054 |
Grade A, and why
recall 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 3d 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.
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.
What it actually says
Brain recall — deep lookup
The user's cross-project knowledge brain is a plain-Markdown vault. Resolve its location first:
$BRAIN_DIR env var if set, else vault_path in ~/.config/claude-brain/config.json, else ~/claude-brain. Call it <vault> below.
If $BRAIN_READ_ALSO is set, it names one or more additional vaults (colon-separated) to search read-only after <vault>. This is how a session that captures into its own vault — a local-model or experimental setup — still gets the benefit of the main brain's history. Never write to a $BRAIN_READ_ALSO vault: no file edits, no new notes, no git commit. Treat it strictly as reference, and say which vault a fact came from when you report it.
Perform a thorough, layered lookup for the topic in $ARGUMENTS (or the current conversation's open question if no arguments). Work down the layers, stopping as soon as the question is fully answered:
- Index: read
<vault>/MEMORY.md— identify which project(s)/topic(s) are relevant. - Pages: read the relevant
<vault>/projects/<name>.mdand/or<vault>/topics/<name>.md. - Session notes: if the page lacks detail, read recent
<vault>/sessions/<project>/*.mdnotes (checkdate:frontmatter; newest first). - Archive (episodic layer): for exact wording, commands, numbers, or anything the notes summarized away, grep the full transcripts:
rg -i "<terms>" <vault>/archive/<project>/— then read the matching section of the hit file. - Cross-project: if the topic spans projects, repeat 2–4 for each; also
rgacross all of<vault>/archive/when unsure where something happened.
Rules:
- Never load whole directories — index first, then only the files the index/grep points to.
- Prefer
status: verifiedcontent overgeneratedwhen they disagree, and say so if they conflict. - Report what you found AND where (file paths), so the user can follow up.
- If nothing is found, say so explicitly and suggest which project's archive is most likely to need a manual look.
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.
- 3d ago First seen · 26 lines · 40 tokens per session scan A 32fa68dc7ed9
recall is a skill published in the GitHub repository vikasgrac/project-brain (2 stars, last pushed 20d ago), licensed MIT. It adds 40 tokens to every session and 540 once invoked, about $0.0002 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 2026-08-31.
Other skills, from other repositories
solomd
Read, search, edit, and version-control any folder of Markdown notes via the SoloMD MCP server. 13 tools including AutoGit per-note history, semantic search, and write-with-sandbox via an accept/reject branch. Includes 11 starter agent recipes (weekly review, todo extract, link suggester, …).
llm-wiki
Build and maintain a persistent, interlinked Obsidian-compatible markdown wiki using Karpathy's LLM Wiki pattern. Extension-backed with auto-generated metadata, guardrails, and 14 custom tools (+3 opt-in agent-trajectory tools).
competitor-feedback
Scan competing markdown editors' user feedback (GitHub issues/discussions/releases closed-source forums) and synthesize themes cross-referenced against SoloMD's own gaps and roadmap. Surfaces unmet demand SoloMD could win on, and features competitors ship that SoloMD lacks. Supports a "subscribe" mode that reports…
my-wiki
Manage local OKF-compatible Markdown My Wiki vaults with an AI agent. Use for capturing webpages, PDFs, Office documents, notes, images, folders, and ZIP bundles as References; maintaining Reference-to-Concept evidence links; searching or answering from a vault; checking or repairing vault health; switching among…
youtube-fetcher
Turn a YouTube video into a structured, Obsidian-ready archival Markdown note containing its transcript, creator metadata, description, chapters, language, and capture provenance. Use when a user shares a YouTube URL or video ID and wants transcripts, captions, subtitles, notes, a knowledge-base record, summarization…
link-memory
Use after important user-approved decisions, when durable context should be proposed or reviewed, and for explicit Link memory lifecycle work: remember, recall, review, update, archive, restore, forget, or explain local memories through the CLI without requiring MCP.