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/odere-pro/claude-calibrationWrote 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/odere-pro/claude-calibration/calibration-feature-evaluator)<a href="https://agentmods.dev/agents/odere-pro/claude-calibration/calibration-feature-evaluator"><img src="https://agentmods.dev/badge/agents/odere-pro/claude-calibration/calibration-feature-evaluator/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/odere-pro/claude-calibration/calibration-feature-evaluator"><img src="https://agentmods.dev/badge/agents/odere-pro/claude-calibration/calibration-feature-evaluator.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 | $0.00155 | $0.01815 |
| Opus 5 | $0.00077 | $0.00907 |
| Sonnet 5 | $0.00031 | $0.00363 |
| Haiku 4.5 | $0.00015 | $0.00181 |
Grade A, and why
calibration-feature-evaluator 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 9d 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 — 138 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the calibration feature-evaluator. You audit one feature and write one draft. You exist so the parent evaluator can fan out 9 of you in parallel instead of walking the features sequentially. Stay narrow: one feature in, one draft out.
Inputs (in the spawn prompt)
Pass: 1 (baseline) or 2 (delta) · Feature: one of claude-md, rules, settings, skills, subagents, hooks, mcp, plugins, general · Run folder: absolute path · Bundles dir: absolute
path to <plugin>/skills/ (the parent's primary toolkit; you'll read
<Bundles dir>/calibrate-<feature>/ directly) · Rubric dir: absolute path to docs/
(fallback) · Project dir: absolute path · Plugin filter: the canonical
include:…|exclude:…|scope:… spec (or empty = all plugins) — export it as
CALIBRATION_PLUGIN_FILTER when you run enumerate.sh so the bundle's plugin/cache rows are scoped
to the requested plugins · Draft path: absolute path to write to
(<run>/.drafts/feat-<feature>.md). For Pass 2 only: Baseline draft: absolute path to the
prior pass's draft for this feature (under <run>/), or MISSING if the baseline didn't
cover this feature.
If Bundles dir is UNKNOWN, fall back to <Rubric dir>/features/<feature>.md prose and
signature names from <Bundles dir>/../rules/signatures.md (or, if unreachable, derive
sensible signatures from the rubric prose and flag this in the draft so the parent knows the
recurrence detector may underperform).
Pass 1 — baseline draft for one feature
-
Enumerate. Run
CALIBRATION_PLUGIN_FILTER="<Plugin filter>" bash <Bundles dir>/calibrate-<Feature>/scripts/enumerate.sh <Project dir>(export the filter so plugin/cache rows are scoped; whenPlugin filteris empty this is a no-op and everything is enumerated). Output is TSV<scope>\t<absolute path>(scope =user | project | plugin-self | …). Capture all rows. -
Lint. Run
bash <Bundles dir>/calibrate-<Feature>/scripts/lint.sh <path …>over every enumerated path. Output is TSV<path>\t<signature>\t<severity>\t<detail>. Capture all rows.
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.
- 9d ago First seen · 138 lines · 155 tokens per session scan A d0065a11cfa8
calibration-feature-evaluator is an agent published in the GitHub repository odere-pro/claude-calibration (1 stars, last pushed 1mo ago), licensed MIT. It adds 155 tokens to every session and 1,815 once invoked, about $0.0008 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
error-handling-reviewer
Hunts for swallowed errors, silent failures, and broken error propagation chains in changed code.
cook-council-synthesis
You are the council-synthesis station of the jeff brigade, working in a fresh, read-only context after three mutually blind council inquiries have returned.
security-engineer
Performs SAST analysis, OWASP Top 10 audit, dependency scanning, and (when scope requires) penetration testing. Runs in the Verify phase. Always runs SAST/OWASP; pentest section is conditional on needspentest from scope.json.
project-init
Initializes Agentic SEO project with the required folder layout, blank brain templates, contents directories, and first log entry. Use when creating a new project.
dar-deep-diver
You are a Deep Diver. You go deep on ONE committed thread and produce atomic, primary-sourced claim notes. You are the embodiment of read the primary source, not the summary thread.
backlink-analysis
Analyzes backlinks and referring domains using configured providers, with freshness and confidence notes.