prompt-optimizer SKILL.md skillA

prompt-optimizer from haabe/mycelium is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

Install it, pinned
mdr add haabe/mycelium/prompt-optimizer@git:20260704.7639f18

Pins git:20260704.7639f18 in mdr.lock. Pin the fingerprint instead.

Audit grade
A, 17 of 17 checks passed
Kind
SKILL.md skill
Repository
haabe/mycelium
Path
plugins/mycelium/skills/prompt-optimizer/SKILL.md
Versions recorded
5
Latest version
git:20260704.7639f18, committed Jul 4, 2026
First version recorded
May 9, 2026
Size
2,186 bytes
Repository stars
51
License
MIT

A/B test CLAUDE.md instruction changes against eval benchmarks. Capture baselines, test variants, compare results.

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

# Prompt Optimizer

Systematically improve Mycelium instructions through measurement. Adapted from n-trax.

## Commands

### `baseline` -- Capture current performance
1. Run `/mycelium:eval-runner run-split optimization` — record as optimization scores
2. Run `/mycelium:eval-runner run-split holdout` — record as holdout scores
3. Record both to `.claude/optimization/baseline.json`: timestamp, CLAUDE.md hash, optimization metrics, holdout metrics, overall and per-category metrics

### `test <variant>` -- Test a variant
1. Read variant from `.claude/optimization/variants/<variant>.md`
2. Apply the CLAUDE.md changes described
3. Run `/mycelium:eval-runner run-split optimization` — this is the hill-climbing signal
4. Run `/mycelium:eval-runner run-split holdout` — this validates generalization
5. Store results in `.claude/optimization/results/<variant>.json`
6. Compare against baseline. Flag **overfitting** if optimization improves but holdout degrades.
7. Do NOT auto-revert -- let user decide

### `report` -- Compare all variants
Generate comparison table with split-aware columns:
```
| Variant | Opt Pass Rate | Holdout Pass Rate | Delta Opt | Delta Holdout | Overfit? | Decision |
…

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint
mdr add haabe/mycelium/prompt-optimizer@sha256:b253c400e961eaf8

The line at the top of this page pins the label git:20260704.7639f18, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_koih44n6vip6jczv.svg)](https://markdownregistry.com/a/art_koih44n6vip6jczv)

5 badge image fetches in 30 days. What the badge shows.

Versions

The registry has recorded 5 versions of this file, between May 9, 2026 and Jul 4, 2026, about one every 14 days.

versioncommittedcommitsizeaudit
git:20260704.7639f18 latest2026-07-04 7639f18 2,186 BA view · diff
git:20260525.9fac9422026-05-25 9fac942 2,186 BA view · diff
git:20260522.1825b222026-05-22 1825b22 1,876 BA view · diff
git:20260509.0d172102026-05-09 0d17210 1,862 BA view · diff
git:20260509.46569102026-05-09 4656910 1,826 BA view

Audit of the latest version

A  17 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (2186 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

haabe/mycelium · 51 stars · license MIT · pushed 2026-11-06 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_koih44n6vip6jczv
GET https://markdownregistry.com/api/v1/resolve?ref=haabe/mycelium/prompt-optimizer
GET https://markdownregistry.com/api/v1/blob/b253c400e961eaf89aca958ee22c389a0de78a74c16703516f6c7f68da7dc8c4

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.

More from haabe/mycelium

AGENTS.md AGENTS.md file
haabe/mycelium · AGENTS.md
git:20260616.1a21799 · audit A · 51 stars
CLAUDE.md CLAUDE.md file
haabe/mycelium · CLAUDE.md
git:20261106.95fe3e6 · audit A · 51 stars
CLAUDE.md@plugins/mycelium/domains/delivery CLAUDE.md file
haabe/mycelium · plugins/mycelium/domains/delivery/CLAUDE.md
git:20260530.bd07760 · audit A · 51 stars
CLAUDE.md@plugins/mycelium/domains/discovery CLAUDE.md file
haabe/mycelium · plugins/mycelium/domains/discovery/CLAUDE.md
git:20260922.f73af47 · audit A · 51 stars
CLAUDE.md@plugins/mycelium/domains/quality CLAUDE.md file
haabe/mycelium · plugins/mycelium/domains/quality/CLAUDE.md
git:20260509.20850da · audit A · 51 stars
a11y-check SKILL.md skill
haabe/mycelium · plugins/mycelium/skills/a11y-check/SKILL.md · Accessibility audit, scoped to the surfaces a product actually has. Detects web / rendered_markdown / terminal /…
git:20260904.e2af329 · audit A · 51 stars
adopt SKILL.md skill
haabe/mycelium · plugins/mycelium/skills/adopt/SKILL.md · Bring Mycelium into a project that already has code. Detects that the repo predates the framework, asks before touching…
git:20261011.6a101df · audit A · 51 stars
assumption-test SKILL.md skill
haabe/mycelium · plugins/mycelium/skills/assumption-test/SKILL.md · Design the smallest viable test to validate or invalidate a critical assumption. Based on Torres's assumption testing…
git:20261024.ac08e78 · audit A · 51 stars
bias-check SKILL.md skill
haabe/mycelium · plugins/mycelium/skills/bias-check/SKILL.md · Use before any research activity or significant decision. Reviews cognitive biases relevant to the current stage.
git:20261007.33e0656 · audit A · 51 stars
bvssh-check SKILL.md skill
haabe/mycelium · plugins/mycelium/skills/bvssh-check/SKILL.md · Use to evaluate whether current work aligns with Better Value Sooner Safer Happier. Run at diamond completion and…
git:20260915.6d95f44 · audit A · 51 stars
canvas-health SKILL.md skill
haabe/mycelium · plugins/mycelium/skills/canvas-health/SKILL.md · Lint canvas files for staleness, missing fields, inconsistent evidence types, and orphaned references. Run periodically…
git:20261106.81ee2f0 · audit A · 51 stars
canvas-sync SKILL.md skill
haabe/mycelium · plugins/mycelium/skills/canvas-sync/SKILL.md · Synchronize canvas state across team sessions via git. Ensures all team members see the same product knowledge.
git:20261003.0391f14 · audit A · 51 stars

Every file in haabe/mycelium

Other files named prompt-optimizer

prompt-optimizer SKILL.md skill
cbrock84/headcount · plugins/technology/skills/prompt-optimizer/SKILL.md · Turns rough intent or a weak prompt into a reliable one — diagnosing why output is inconsistent, restructuring the…
git:20260901.aa45383 · audit A · 2,543 stars
prompt-optimizer SKILL.md skill
daymade/claude-code-skills · prompt-optimizer/SKILL.md · Transforms vague prompts or feature requests into precise, testable specifications using EARS (Easy Approach to…
git:20260923.943df48 · audit A · 1,519 stars
prompt-optimizer SKILL.md skill
testany-io/testany-agent-skills · plugins/testany-llm/skills/prompt-optimizer/SKILL.md · 撰写提示词、优化提示词、改写 Prompt、Prompt 优化。Use when: 用户需要创建或修改可执行提示词;按任务、调用方输出契约与实际工具能力优化,不按模型品牌强制格式。
git:20260914.a55b8c5 · audit A · 82 stars
prompt-optimizer SKILL.md skill
frankxai/agentic-creator-os · .claude/skills/prompt-optimizer/SKILL.md · Analyze raw prompts, identify intent and gaps, match ECC components (skills/commands/agents/hooks), and output a…
v1.0.0 · audit A · 12 stars
prompt-optimizer SKILL.md skill
softmisst/tudo-sobre-vibe-coding · Claude Code/Claude Skills/prompt-optimizer/prompt-optimizer/SKILL.md · Analyze raw prompts, identify intent and gaps, match ECC components (skills/commands/agents/hooks), and output a…
v1.0.0 · audit A · 8 stars

Browse by kind, by grade A, or by owner.