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/verve-sh/session-recall/claude-mdgit clone --depth 1 https://github.com/verve-sh/session-recallWrote 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/verve-sh/session-recall/claude-md)<a href="https://agentmods.dev/instructions/verve-sh/session-recall/claude-md"><img src="https://agentmods.dev/badge/instructions/verve-sh/session-recall/claude-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 5.1 | $0.01209 | $0.01209 |
| Opus 5 | $0.00605 | $0.00605 |
| Sonnet 5 | $0.00242 | $0.00242 |
| Haiku 4.5 | $0.00121 | $0.00121 |
Grade A, and why
session-recall CLAUDE.md 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 5d 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.
How it starts
The opening of the file, as written. The whole thing — 82 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Session Recall Plugin
Standalone Claude Code plugin: hybrid FTS5 trigram + vector cosine search over past sessions, fused via RRF with recency boost.
Commands
cd plugins/session-recall && npm install --legacy-peer-deps
cd plugins/session-recall && npm run build
cd plugins/session-recall && npm run test
Architecture
src/
index.ts # MCP server setup, tool registration, stdio transport
tools.ts # Zod schemas + handlers for session_search, session_drill, session_health
lib.ts # Core: indexSession, search (hybrid FTS5 + vec0 cosine, RRF fusion), drill, health
db.ts # SQLite singleton: better-sqlite3 + sqlite-vec, sessions DB at ~/.claude/sessions-index.db
marker.ts # SessionEnd hook: zero native deps, marks session transcript for async indexing
cli.ts # CLI entry: session-recall index|search|drill|health
hooks/
hooks.json # SessionEnd hook definition
session-end-hook.sh # Hook script — invokes CLI to index completed session
scripts/
start-mcp.sh # MCP launcher — builds on first run, then execs the server
Build model
plugin.json launches scripts/start-mcp.sh, not node dist/index.js. The launcher
execs the server straight through when dist/index.js exists, and otherwise runs
npm install --legacy-peer-deps && npm run build first.
dist/ stays gitignored and cannot be committed. better-sqlite3 loads a
platform-specific better_sqlite3.node and sqlite-vec ships a per-platform vec0
extension, so build output is only valid on the machine that produced it — a prebuilt
artifact would break every install on a different OS or arch. (dispatch-lens has no
native deps, so it commits a bundle instead.)
Three invariants when touching the launcher:
- Nothing may write to stdout. stdout is the JSON-RPC transport; npm writes
progress there by default, which is why every build command is redirected
>&2. - Concurrent starts are serialised through an atomic
mkdir "$ROOT/.build.lock". Several Claude Code sessions can launch the server at once, and twonpm installruns in one directory corruptnode_modules. Release the lock beforeexec—execreplaces the process, so theEXITtrap never fires. - Readiness means "a database opens", not "
dist/index.jsexists". Claude Code runs its ownnpm installat install time with install scripts disabled, sonode_modules/better-sqlite3arrives with nobuild/andbetter_sqlite3.nodeis never compiled. A laternpm installdoes nothing about it — the tree already satisfies the lockfile — so the launcher mustnpm rebuild better-sqlite3explicitly. Left unbuilt, the server completes the MCP handshake and lists its tools, then fails every call with "Internal error".
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.
- 5d ago First seen · 82 lines · 1,209 tokens per session scan A 1868adea1ebd
session-recall CLAUDE.md is an instructions file published in the GitHub repository verve-sh/session-recall (1 stars, last pushed 22d ago), licensed Apache-2.0. It adds 1,209 tokens to every session, about $0.0060 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 instructions, from other repositories
aeon CLAUDE.md
Claude Code instructions for aeonfun/aeon, covering aeon, how aeon works, strategy, voice and soul file hierarchy (read in this order).
cognirepo CLAUDE.md
Claude Code instructions for ashlesh-t/cognirepo, covering claude.md, key rules, session start sequence (run in this order), behavioral confirmation rule and personas (cognirepo-402, cognirepo-403).
wayland-core copilot-instructions.md
Copilot instructions for FerroxLabs/wayland-core, covering ijfw rules, output discipline, memory routing, context discipline and cross-audit.
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.
RNR-Enhanced-Cognee AGENTS.md
AGENTS.md instructions for vincentspereira/RNR-Enhanced-Cognee, covering rnr enhanced cognee implementation for codex, critical requirements, 1. ascii-only output (no unicode encoding), 2. dynamic categories (no hardcoded categories) and 3. standard memory mcp interface.