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/ramarahmanda/tsubasaWrote 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/ramarahmanda/tsubasa/plugin)<a href="https://agentmods.dev/plugins/ramarahmanda/tsubasa/plugin"><img src="https://agentmods.dev/badge/plugins/ramarahmanda/tsubasa/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/ramarahmanda/tsubasa/plugin"><img src="https://agentmods.dev/badge/plugins/ramarahmanda/tsubasa/plugin.svg" alt="Reviewed on agentmods" width="80" height="20"></a>Grade A, and why
tsubasa 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 1mo 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": "tsubasa",
"description": "Captain persona: repo-resident domain expert with a git-stored knowledge graph, ambient knowledge capture, and ADR-threaded delegation.",
"version": "0.1.20",
"author": { "name": "Rama Rahmanda" },
"homepage": "https://github.com/ramarahmanda/tsubasa"
}
What it installs
The manifest is a name and a version. 5 skills, 4 hooks travel with it, and installing the plugin installs all of them — 372 tokens a session between them. Each is measured on its own page, and each can be installed alone.
- Skill captain-recall A 65 tokens
- Skill captain-delegate A 54 tokens
- Skill captain-onboard A 110 tokens
- Skill captain-inject A 85 tokens
- Skill captain-capture A 58 tokens
- Hook SubagentStart A not measured
- Hook PreToolUse A not measured
- Hook UserPromptSubmit A not measured
- Hook SessionStart 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.
- 1mo ago First seen · 8 lines scan A 86b7ecf84df7
tsubasa is a plugin published in the GitHub repository ramarahmanda/tsubasa (9 stars, last pushed 1mo 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 2026-08-31.
Other plugins, from other repositories
memsearch
Automatic semantic memory for Claude Code — remembers what you worked on across sessions.
evolve-lite
Recall, save, and share reusable Evolve entities.
autograph
Schema-as-code memory engine for Obsidian vaults that AI agents write to: typed cards, entity dedup, link repair, update-in-place (supersede + History), Ebbinghaus decay, MOC generation, and a vault health score. Plain Markdown you own, Python stdlib only. Ships the autograph skill plus a /research slash command that…
cc-memory
Persistent memory for Claude Code — automatic save/restore across compactions and sessions. Anti-patch reconcile-on-write + LLM-judged semantic de-duplication, forced PROGRESS.md handoff with per-session annotation, live PLAN.md anchor, injection observability, FTS5 search, AI-judged extraction with Haiku + local…
presence
Continuous-collaboration layer for Claude Code: living project model, outcome telemetry, event digest, and calibrated confidence. Fully local, install-once, applies to every project.
jarvis-os
A persistent memory hub for Claude Code that loads into every session from any directory, plus a curated set of research and workflow skills.