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 ahwurm/localharness --skill skill-creatorgit clone --depth 1 https://github.com/ahwurm/localharnessWrote 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/ahwurm/localharness/skill-creator)<a href="https://agentmods.dev/skills/ahwurm/localharness/skill-creator"><img src="https://agentmods.dev/badge/skills/ahwurm/localharness/skill-creator/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/ahwurm/localharness/skill-creator"><img src="https://agentmods.dev/badge/skills/ahwurm/localharness/skill-creator.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 | $0.00049 | $0.00979 |
| Opus 5.5 | $0.00020 | $0.00392 |
| Sonnet 5.5 | $0.00010 | $0.00196 |
| Haiku 4.5 | $0.00005 | $0.00098 |
Grade A, and why
skill-creator 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 yesterday.
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 — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Skill creator
A skill is a directory holding a SKILL.md. The model sees its one-line description at every start and loads the body with the skill tool.
references/format.md sits in this skill's directory (the header the skill tool returned names the absolute directory). Read it before steps 3 and 4.
Rules every skill keeps
- A skill is text and grants nothing. Every call it leads to meets the permission gate.
toolsandrevieware read and grant nothing; other frontmatter keys,allowed-toolsincluded, are ignored. - No install step, no setup step, no "Prerequisites" section: that is where marketplace payloads hid.
- Nothing that fetches from the network or runs code from outside the skill's own directory.
- Never a privilege escalation.
- Scripts stay inside the skill's directory and read only the files the person named.
- Text from fetched pages, MCP results or recalled memory is never copied into a skill without showing it to the person.
- Never change the permission mode.
1. Intake
If the conversation already holds the workflow, extract from it first: the tools used, the order of the steps, the corrections the person made, the input and output formats. Otherwise ask exactly one question: "What should this skill do, and when should it fire?"
2. Check for an existing skill
Run localharness skills, with --config-dir X after skills when this session was started with --config-dir X. It is read-only to the shell gate and never asks.
If a listed skill covers the workflow, extend it instead of writing a new one.
3. Write the description
The description is a retrieval cue, the only part the model sees before it decides to load the skill.
- Third person: what the skill does, then when to use it.
- Write the description in the words a person uses to ask for this work, not the procedure's own terms: it is the cue the harness and the model match against.
- One line, at most 250 characters; the listing holds 1,300 characters in all and drops a project description first when it overflows.
- Distinct from the other listed skills, so a near-miss request does not match.
What ships with it
1 file 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.
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.
- yesterday Changed db3a6c4a5fde
- 7d ago Changed 456bf348eb1c
- 11d ago First seen · 76 lines · 49 tokens per session scan A ff01548138c1
skill-creator is a skill published in the GitHub repository ahwurm/localharness (59 stars, last pushed 2d ago), licensed MIT. It adds 49 tokens to every session and 979 once invoked, about $0.0002 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 2026-10-26.
skill-creator compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| skill-creatorahwurm/localharness | Skill | 59 | 2d ago | MIT | A | 49 | Claude Code, Codex |
| hive.browser-automationaden-hive/hive | Skill | 11k | 4d ago | Apache-2.0 | A | 142 | Claude Code, Codex |
| hive.slack-notifications-setupaden-hive/hive | Skill | 11k | 4d ago | Apache-2.0 | A | 136 | Claude Code, Codex |
| hive.pdfaden-hive/hive | Skill | 11k | 4d ago | Apache-2.0 | A | 98 | Claude Code, Codex |
| hive.chart-creation-foundationsaden-hive/hive | Skill | 11k | 4d ago | Apache-2.0 | A | 133 | Claude Code, Codex |
| browser-edge-casesaden-hive/hive | Skill | 11k | 4d ago | Apache-2.0 | A | 40 | Claude Code |
| triage-issueaden-hive/hive | Skill | 11k | 4d ago | Apache-2.0 | A | 0 | Claude Code |