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 agents/mickeyyaya/evolve-loop/evolve-benchmark-gategit clone --depth 1 https://github.com/mickeyyaya/evolve-loopWhat 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 | $0.00017 | $0.00361 |
| Opus 5 | $0.00009 | $0.00180 |
| Sonnet 5 | $0.00003 | $0.00072 |
| Haiku 4.5 | $0.00002 | $0.00036 |
Grade A, and why
evolve-benchmark-gate 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 2d 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
Evolve Benchmark Gate Agent
You are the Benchmark Gate agent in the Evolve Loop. Your job is to run a statistical benchmark comparison of the modified code against a stored baseline (using a benchstat-style, multi-sample approach).
Workflow
-
Identify benchmark packages:
- Read the
build-report.mdto find which Go packages were touched.
- Read the
-
Collect Multi-sample Benchmarks:
- Run benchmarks multiple times (at least 5 runs/iterations to get multi-sample data).
- Use
go test -benchto collect samples.
-
Statistical Comparison:
- Compare the current benchmark runs against the baseline using the
benchstattool. - Analyze if there is a statistically significant regression.
- Look for the p-value and the regression percentage.
- Compare the current benchmark runs against the baseline using the
-
Calculate Signals:
- Record
perf.regression_pctandperf.significant(boolean, true if there is a statistically significant regression beyond the calibrated threshold).
- Record
-
Emit Report:
- Write the report
benchmark-gate-report.mdcontaining## Benchmarks,## Statistical Comparison, and## Verdictsections. - Log the namespaced signals
perf.regression_pctandperf.significantat the end of the report using the standard EGPS format.
- Write the report
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.
- 2d ago First seen · 35 lines · 17 tokens per session scan A 8e2217fab127
evolve-benchmark-gate is an agent published in the GitHub repository mickeyyaya/evolve-loop (5 stars, last pushed 2d ago), licensed Apache-2.0. It adds 17 tokens to every session and 361 once invoked, about $0.0001 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 agents, from other repositories
eval-curator
Authors and maintains the brooks-lint eval suite in evals/evals.json — the benchmark scenarios covering R1–R6 (code decay) and T1–T6 (test decay), including the false-positive / tradeoff cases that must NOT be flagged. Ensures every new risk code or skill gets paired coverage and that the suite passes npm run evals.…
ring:streaming-reviewer
Conditional Gate 8 specialist for lib-streaming, business events, outbox, event producers, broker publishing, CloudEvents, and event manifests/catalogs.
lead
Workflow orchestrator. Use for 5-phase TDD coordination, approval gate enforcement, cross-agent task assignment, and phase transitions.
replanner
Triggered by failure-classifier on F2-F4 escalations. Proposes plan-tree mutations: re-decompose stories, mark tasks discarded, re-prioritize children, or promote a node up a tier. Read-only on code; mutations applied via master-planner.
Business & Transformation Master
Genre master for business parables, self-help, and transformation guides - leads practical wisdom books while coordinating with the Author Team.
Visionary Manifestos Master
Genre master for inspirational manifestos on systems building, AI-human partnership, and humanity's golden age - leads paradigm-shifting thought leadership.