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.
git clone --depth 1 https://github.com/joseairosa/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/plugins/joseairosa/recall/claude-plugin)<a href="https://agentmods.dev/plugins/joseairosa/recall/claude-plugin"><img src="https://agentmods.dev/badge/plugins/joseairosa/recall/claude-plugin/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/plugins/joseairosa/recall/claude-plugin"><img src="https://agentmods.dev/badge/plugins/joseairosa/recall/claude-plugin.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 12d 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
{
"name": "recall",
"version": "1.18.8",
"description": "Recall Memory-as-a-Service — persistent memory, semantic search, and context management across Claude Code sessions. Connects to recallmcp.com or self-hosted instances.",
"author": {
"name": "Jose Airosa",
"email": "[email protected]",
"url": "https://github.com/joseairosa"
},
"homepage": "https://recallmcp.com",
"repository": "https://github.com/joseairosa/recall",
"license": "MIT",
"keywords": ["memory", "recall", "mcp", "persistence", "semantic-search", "context", "rlm"]
}
What it installs
The manifest is a name and a version. 7 commands, 3 agents, 6 hooks travel with it, and installing the plugin installs all of them — 142 tokens a session between them. Each is measured on its own page, and each can be installed alone.
- Command decompose A 16 tokens
- Command rlm-status A 15 tokens
- Command load-context A 18 tokens
- Command recall-loop A 36 tokens
- Command recall-set-key C 0 tokens
- Command recall-update C 0 tokens
- Command setup D 0 tokens
- Agent result-aggregator A 19 tokens
- Agent task-decomposer A 18 tokens
- Agent context-loader A 20 tokens
- Hook SessionStart A not measured
- Hook Stop A not measured
- Hook PostToolUse A not measured
- Hook SessionEnd A not measured
- Hook PreCompact A not measured
- Hook PostToolUseFailure A not measured
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.
- 12d ago Changed 61f40ff6945a
- 13d ago Changed ab2f82afdcc6
- 16d ago Changed b5fd37ee9ce3
- 1mo ago First seen · 15 lines scan A 8c7b5b70387e
recall is a plugin published in the GitHub repository joseairosa/recall (176 stars, last pushed 13d ago), licensed MIT. Its token cost is not measured: this kind of file is read by the harness, not the model. 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 plugins, from other repositories
yantrikdb
Cognitive memory for AI agents — semantic recall, a knowledge graph, contradiction detection and procedural learning. Local-first: one SQLite file, no account, no telemetry.
cairn
AI Context Continuity Engine — warns before the context window degrades, hands off a faithful 7-bucket brief, and keeps a decision/evidence ledger across sessions and models.
vibebox
Local-first structured memory middleware for Claude Code and other AI coding agents.
mori
Persistent, cross-device memory for your AI coding agents — self-hosted and AGPL, so your knowledge never leaves your own infrastructure. Connect to a Mori server you run; every session distils into durable, searchable memory, so agents start informed, not cold. /brief loads context, /pensieve searches it.
distil
Live session-first savings status line, /distil commands (stats, dashboard, shadow equivalence, doctor, onboard, trajectory certificate, savings badge) for distil — certified context compression.
smartcompact
Human-in-the-loop context compaction for Claude Code. You choose which turns survive; a SessionStart hook re-injects them deterministically after every compact and resume.