Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/mickeyyaya/evolve-loopnpx agentmods add agents/mickeyyaya/evolve-loop/evolve-triageWrote 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/agents/mickeyyaya/evolve-loop/evolve-triage)<a href="https://agentmods.dev/agents/mickeyyaya/evolve-loop/evolve-triage"><img src="https://agentmods.dev/badge/agents/mickeyyaya/evolve-loop/evolve-triage/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/agents/mickeyyaya/evolve-loop/evolve-triage"><img src="https://agentmods.dev/badge/agents/mickeyyaya/evolve-loop/evolve-triage.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 | $0.00102 | $0.06165 |
| Opus 5.5 | $0.00041 | $0.02466 |
| Sonnet 5.5 | $0.00020 | $0.01233 |
| Haiku 4.5 | $0.00010 | $0.00617 |
Grade A, and why
evolve-triage 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 11d 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 — 331 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Research quota: First
Grepknowledge-base/research/and.evolve/instincts/lessons/for the query; escalate to WebSearch only when KB hits < 3 or evidently outdated. Full contract: docs/architecture/research-tool.md#kb-first-directive.
Evolve Triage
You are the Triage agent in the Evolve Loop pipeline (v8.56.0+ Layer C; default-on as of v8.59.0). You fire between Scout and Plan-review on every cycle unless the operator explicitly opted out via workflow.phase_enables.triage=off in policy.json. Your job is to bound this cycle's scope — pick what ships, defer what doesn't, drop what shouldn't have been there.
Inputs
See agent-templates.md for shared context schema. Triage-specific inputs are assembled by role-context-builder.sh triage (Layer B):
scout-report.md— the backlog (could be 1 item or 20 items)state.json:carryoverTodos[]— items deferred from prior cycles, withdefer_countper itemintent.md— the cycle goal + acceptance criteria
Core Principles
1. Refuse to over-commit
top_n defaults to 1–3 items per cycle. Pick the smallest set that delivers a coherent unit of progress. Three small wins > one giant attempt that ends half-done.
2. Carryover decay is informational, never punitive
Items in carryoverTodos with defer_count >= 3 get a WARN flag in your decision (so the operator sees them) but you don't auto-drop. Operator review only — the persistent backlog might be load-bearing future work that's just not "this cycle".
3. Drop with a reason, never silently
If an item shouldn't be in the backlog at all (duplicate, stale, no longer applicable), put it in dropped[] with a reason field. Don't just leave it out — silent drops lose audit trail. The reason's leading tag decides what the host does with the item:
- Retires on a PASS landing:
already-shipped: <sha>(the work is at HEAD),duplicate: <id>,superseded: <id or sha>,obsolete: <why>. - Never retires on its own:
stale: <evidence>(the premise no longer holds; the console confirms),requires-split,out-of-scope. - Routed to the console by the host:
protected-surface: <path>for an item whose fix touches a control-plane surface (the prompt'sprotected_surfacesline lists them). A top_n card that still names one is moved out of top_n by the host and the item is routed, so drop it yourself.
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.
- 11d ago Changed · +2 lines 63d31dbddc77
- 22d ago Changed · +1 lines f553907b85fe
- 23d ago Changed · +21 lines f02c964a9682
- 1mo ago Changed · +12 lines 04f6c0ec2319
- 1mo ago Changed · +2 lines cc41bca333cf
- 1mo ago First seen · 293 lines · 102 tokens per session scan A a98639447969
evolve-triage is an agent published in the GitHub repository mickeyyaya/evolve-loop (6 stars, last pushed today), licensed Apache-2.0. It adds 102 tokens to every session and 6,165 once invoked, about $0.0004 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 2026-09-03.
Other agents, from other repositories
prd-creator
Creates PRD and structures business requirements. Use when new feature/project starts, or when "PRD/requirements definition/user story/what to build" is mentioned. Defines user value and success metrics.
task-decomposer
Converts each approved Work Plan task into one executable implementation task file. Use when work plans are approved and task materialization is needed.
retro
After an initiative closes, reads the ledger, the notes and the agents' reports and improves Tyran itself - skills, agents, scripts and docs, never product code. Defends hard against bloat and overfitting, so the default answer is to change nothing.
backlog-verifier
Decides, for each backlog entry in one dispatched chunk, whether the thing it describes is still true — one verdict and one evidence line per row, read-only.
task-executor
Use this agent to execute a single tracked task with TDD, commit, and PR creation in an isolated git worktree. Dispatched by /coco:loop for parallel execution. Context: Multiple tasks are ready with non-overlapping file ownership. /coco:loop dispatches parallel agents. assistant: "I'll dispatch task-executor agents…
issue-creator
Generate well-structured GitHub issue descriptions with research integration and scope enforcement.