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/albiol2004/trio-agent-loopWrote 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/commands/albiol2004/trio-agent-loop/trio-bridge)<a href="https://agentmods.dev/commands/albiol2004/trio-agent-loop/trio-bridge"><img src="https://agentmods.dev/badge/commands/albiol2004/trio-agent-loop/trio-bridge/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/commands/albiol2004/trio-agent-loop/trio-bridge"><img src="https://agentmods.dev/badge/commands/albiol2004/trio-agent-loop/trio-bridge.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00018 | $0.01238 |
| Opus 5.5 | $0.00007 | $0.00495 |
| Sonnet 5.5 | $0.00004 | $0.00248 |
| Haiku 4.5 | $0.00002 | $0.00124 |
Grade A, and why
trio-bridge 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.
How it starts
The opening of the file, as written. The whole thing — 86 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the orchestrator of a read-only background scan. You do no scanning or editing yourself — you dispatch scout subagents to find candidate work, dedupe their findings, write proposal cards, and report a short summary. One invocation = one scan pass; re-running is safe (idempotent). You never start loops and never edit project code.
This command runs with the project root as its working directory — every path below is relative to it.
0. Preflight
- Create
proposals/at the project root if it does not exist. - List existing cards:
proposals/*.md. Their frontmatterfingerprintvalues are the dedupe set. - List live missions: every
loop*/GOAL.mdwith amission:line. Work already claimed by a live mission is out of scope.
1. Dispatch scouts (background, parallel)
Dispatch 1-3 read-only trio-scout subagents as background tasks, in parallel, each with a disjoint area or lens (e.g., docs; manifests/checks; code hygiene). The task item MUST name the trio-scout agent and MUST run in the background so nothing blocks on them. Prompt each scout with:
- a READ-ONLY mandate (inspect only; never edit, never run state-changing commands),
- its scan area, and
- the candidate categories to look for: stale or contradictory documentation; manifest/config drift; failing or missing checks; TODO/FIXME clusters; missing verification for recent changes,
- the return contract: a concise list of candidate objects, exactly
{origin, summary, evidence, confidence, class}:origin: the file/path the finding points at,summary: one sentence naming the concrete problem (deterministic wording — it feeds the fingerprint),evidence:file:linereferences or the exact failing command,confidence: low | medium | high,class: A | B | C (definitions in step 4),
- no prose and no file contents in the reply — only the list.
Scout results auto-deliver in the background; never busy-poll and never block your turn waiting for them.
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 · 86 lines · 18 tokens per session scan A eb498cbaeb70
trio-bridge is a command published in the GitHub repository albiol2004/trio-agent-loop (2 stars, last pushed 12d ago), licensed MIT. It adds 18 tokens to every session and 1,238 once invoked, about $0.0001 per session on Opus 5.5. 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 commands, from other repositories
arrow-maintenance
Run an audit-and-update pass on the docs/arrows/ overlay. Dispatches on project state — audits an existing overlay, bootstraps the overlay from existing LID docs, or redirects to /map-codebase or /linked-intent-dev for projects without LID.
differential-audit
Audit EARS↔code coherence via a bidirectional differential. With no arguments, opens a scoping conversation that interprets what the user wants audited, maps it to arrows, and captures an explicit EARS selection. With one or more EARS IDs, audits those specs directly using configured defaults. Requires docs/arrows/…
map-codebase
Bootstrap LID in an existing (brownfield) codebase. Reads every file in scope, proposes lens-based clusterings, walks the user through reconciliation, generates skeleton LLDs/HLD/EARS and the arrow overlay, then prompts the user to flesh out the skeletons. Token-intensive by design.
commit
A command that examines staged Git changes and proposes a commit message using the Conventional Commits format, such as feat, fix, or docs. Git is a tool for tracking changes to code.
crear-skill
Crear, probar y mejorar skills de forma iterativa. Usa cuando el usuario dice "crear skill", "crear habilidad", "build skill", "create a skill", "skill development", "desarrollar skill", "nueva habilidad". Usa el workflow: draft → test → review → improve → repeat con evaluación cuantitativa y qualitative review.
claude-request
Send a prompt to Claude Code via the LLM gateway with session tracking.