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 skills add yungbose/upskill --skill optimise-skillgit clone --depth 1 https://github.com/yungbose/upskillWrote 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/skills/yungbose/upskill/optimise-skill)<a href="https://agentmods.dev/skills/yungbose/upskill/optimise-skill"><img src="https://agentmods.dev/badge/skills/yungbose/upskill/optimise-skill/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/skills/yungbose/upskill/optimise-skill"><img src="https://agentmods.dev/badge/skills/yungbose/upskill/optimise-skill.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 | $0.00110 | $0.03570 |
| Opus 5.5 | $0.00044 | $0.01428 |
| Sonnet 5.5 | $0.00022 | $0.00714 |
| Haiku 4.5 | $0.00011 | $0.00357 |
Grade A, and why
optimise-skill 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 2mo 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 — 221 lines — stays where its author put it; the contents beside it link to each section on GitHub.
optimise-skill
A decision-policy optimiser. It tunes ONE judge-shaped document (a rubric that
makes a recurring, scorable decision) against an outcome-labelled corpus, accepting
an edited rubric only when a held-out score strictly improves and adversarial gates
pass. It adopts microsoft/SkillOpt (on OPRO foundations) as its optimiser_engine,
vendored unedited under vendor/skillopt/; everything skill-specific lives in the
rails outside it.
Announce at start: "I'm using optimise-skill to tune <target> against its corpus (autonomy: <preset>)."
It is behavioural-only. Tuning a skill's description: so Claude invokes it is a
different job (different artifact, signal, lifecycle) and stays with the official
skill-creator. See CHARTER.md for the full design and dev/papers.md for the
SkillOpt/OPRO lineage.
Requirements
Run these from the optimise-skill/ directory:
- Python 3.10+ (tested on 3.13) with PyYAML —
pip install -r requirements.txt. PyYAML is the only runtime dependency; the vendored SkillOpt's heavier backends (openai, azure, ray, …) are optional and unused on the default path. - Live tuning picks a provider via the manifest's
model:field:model: claude(default) — needs theclaudeCLI + a Claude Max subscription (shells out toclaude -p; billed against Max, not the metered API).model: openai— for users without Max:pip install openaiand setOPENAI_API_KEY(pay-as-you-go OpenAI billing; Azure OpenAI works similarly via SkillOpt's backend). Note: there is no Anthropic-HTTP-API backend, so using Claude here means the CLI/subscription, not an Anthropic API key.
- Tests:
pip install pytest && python3 -m pytest tests/— runs with stub judges, no LLM required.
When to use — the durable decision rule
Apply this to any signal, any skill, before making it a target:
Is it a recurring decision made by ONE tunable text document, where each decision is scorable and you can counterfactually imagine the edited document scoring better, with enough volume to hold out?
Yes → an optimise-skill target. No → human-triage (one-off fixes, via a judge-shaped edit-queue) or monitoring (metrics).
What ships with it
60 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- CHARTER.md 13 KB
- dev/papers.md 6.9 KB
- references/contract.md 15 KB
- requirements.txt 591 B
- scripts/adapter.py 10 KB runs code
- scripts/attribute.py 10 KB runs code
- scripts/claude_code_judge.py 5.4 KB runs code
- scripts/decision_record.py 6.9 KB runs code
- scripts/default_guards.py 1.1 KB runs code
- scripts/default_scorer.py 2.0 KB runs code
- scripts/gates.py 3.6 KB runs code
- scripts/ledger.py 3.0 KB runs code
- scripts/md_corpus.py 6.5 KB runs code
- scripts/optimise.py 26 KB runs code
- scripts/run_health.py 1.8 KB runs code
- scripts/run.py 4.4 KB runs code
- scripts/seed.py 6.0 KB runs code
- scripts/sentinel.py 1.9 KB runs code
- scripts/setup.py 9.4 KB runs code
- scripts/stratify.py 2.6 KB runs code
- scripts/stuck_events.py 4.7 KB runs code
- scripts/subscription.py 6.9 KB runs code
- scripts/trust.py 1.9 KB runs code
- scripts/validate_manifest.py 6.6 KB runs code
- tests/conftest.py 655 B runs code
- tests/fixtures/escalation_cap_guards.py 790 B runs code
- tests/test_adapter.py 10 KB runs code
- tests/test_attribute.py 7.6 KB runs code
- tests/test_claude_code_judge.py 3.9 KB runs code
- tests/test_decision_record.py 6.6 KB runs code
- tests/test_default_scorer.py 2.6 KB runs code
- tests/test_gates.py 2.9 KB runs code
- tests/test_ledger.py 2.5 KB runs code
- tests/test_md_corpus.py 4.8 KB runs code
- tests/test_optimise.py 22 KB runs code
- tests/test_run_health.py 1.7 KB runs code
- tests/test_run.py 3.3 KB runs code
- tests/test_seed.py 3.6 KB runs code
- tests/test_sentinel.py 1.7 KB runs code
- tests/test_setup.py 4.0 KB runs code
- tests/test_stratify.py 2.7 KB runs code
- tests/test_stuck_events.py 2.6 KB runs code
- tests/test_subscription.py 6.2 KB runs code
- tests/test_trust.py 2.8 KB runs code
- tests/test_validate_manifest.py 6.2 KB runs code
- vendor/skillopt/__init__.py 728 B runs code
- vendor/skillopt/config.py 11 KB runs code
- vendor/skillopt/datasets/__init__.py 324 B runs code
- vendor/skillopt/datasets/base.py 18 KB runs code
- vendor/skillopt/engine/__init__.py 303 B runs code
- vendor/skillopt/engine/trainer.py 90 KB runs code
- vendor/skillopt/envs/__init__.py 36 B runs code
- vendor/skillopt/envs/_template/config_template.yaml 2.3 KB
- vendor/skillopt/envs/_template/env_template.py 3.1 KB runs code
- vendor/skillopt/envs/_template/loader_template.py 3.3 KB runs code
- vendor/skillopt/envs/_template/README.md 616 B
- vendor/skillopt/envs/alfworld/__init__.py 139 B runs code
- vendor/skillopt/envs/alfworld/adapter.py 16 KB runs code
- vendor/skillopt/envs/alfworld/dataloader.py 3.8 KB runs code
- vendor/skillopt/envs/alfworld/prompts/analyst_error.md 2.9 KB
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.
- 2mo ago First seen · 221 lines · 110 tokens per session scan A ff0230a60c80
optimise-skill is a skill published in the GitHub repository yungbose/upskill (14 stars, last pushed 5mo ago), licensed MIT. It adds 110 tokens to every session and 3,570 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 2026-08-30.
optimise-skill compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| optimise-skillyungbose/upskill | Skill | 14 | 5mo ago | MIT | A | 110 | Claude Code, Codex |
| docxsiddharthrath1999/Claude-Opus-4.7 | Skill | 2 | 6mo ago | MIT | A | 168 | Claude Code, Codex |
| claude-apisiddharthrath1999/Claude-Opus-4.7 | Skill | 2 | 6mo ago | MIT | A | 103 | Claude Code, Codex |
| pptxsiddharthrath1999/Claude-Opus-4.7 | Skill | 2 | 6mo ago | MIT | A | 152 | Claude Code, Codex |
| skill-creatorsiddharthrath1999/Claude-Opus-4.7 | Skill | 2 | 6mo ago | MIT | A | 64 | Claude Code, Codex |
| xlsxsiddharthrath1999/Claude-Opus-4.7 | Skill | 2 | 6mo ago | MIT | A | 201 | Claude Code, Codex |
| mcp-buildersiddharthrath1999/Claude-Opus-4.7 | Skill | 2 | 6mo ago | MIT | A | 61 | Claude Code, Codex |