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 Nolane-x/forge-os --skill forgeos-skill-promotiongit clone --depth 1 https://github.com/Nolane-x/forge-osWrote 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/nolane-x/forge-os/forgeos-skill-promotion)<a href="https://agentmods.dev/skills/nolane-x/forge-os/forgeos-skill-promotion"><img src="https://agentmods.dev/badge/skills/nolane-x/forge-os/forgeos-skill-promotion/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/nolane-x/forge-os/forgeos-skill-promotion"><img src="https://agentmods.dev/badge/skills/nolane-x/forge-os/forgeos-skill-promotion.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 | $0.00025 | $0.00340 |
| Opus 5 | $0.00013 | $0.00170 |
| Sonnet 5 | $0.00005 | $0.00068 |
| Haiku 4.5 | $0.00003 | $0.00034 |
Grade A, and why
forgeos-skill-promotion 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.
What it actually says
ForgeOS Skill Promotion
Promotion is an evidence decision, not a popularity contest.
Benchmark design
- Define at least five representative tasks from the target capability and two adversarial or edge-case tasks.
- Run the same model, tools, context budget, and seeds without the skill and with the skill.
- Measure task success, unnecessary tool calls, latency, token usage, policy violations, fabricated claims, and recoverability after failure.
- Repeat enough runs to distinguish a consistent improvement from random variation.
- Test coexistence with neighboring skills to detect retrieval collisions and contradictory instructions.
Promotion gates
Promote only when all conditions hold:
- Task success improves or remains equal while reducing cost or risk.
- No new critical safety or permission failure appears.
- The description retrieves the skill for relevant tasks and avoids irrelevant tasks.
- License and provenance are recorded.
- Required tools exist and expose capability-honest interfaces.
- The skill has an owner, freshness window, revocation path, and regression tests.
ForgeOS state
Use the progression candidate -> evaluated -> stable. A failed regression returns the skill to candidate or disables it. Never overwrite the prior signed record; create a new version linked to the old one.
Verification
Attach benchmark inputs, outputs, verifier results, environment fingerprint, source snapshot, and promotion decision to the catalog entry.
Failure handling
If the skill has no measurable benefit, keep it out of the stable catalog even when its Markdown appears polished.
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 · 40 lines · 25 tokens per session scan A 3e063fc3817d
forgeos-skill-promotion is a skill published in the GitHub repository Nolane-x/forge-os (10 stars, last pushed 1mo ago), licensed MIT. It adds 25 tokens to every session and 340 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 skills, from other repositories
cinematic-video-gen
A prompt-writing guide for making AI-generated videos look more natural and expressive. It covers facial expressions, body movements, and keeping a character consistent between clips.
clanker
Plan and review ERC20 deployment with the Clanker SDK. Use for token configuration, vesting, airdrops, rewards, metadata, and transaction preparation on supported EVM chains.
content-typography
A Chinese-language typography and layout guide for content images such as social-media covers, article headers, and posters. It covers headline wording, line breaks, hierarchy, font weight, alignment, spacing, and platform fit.
article-to-infographic
A planning tool that turns a long article or text into three to five visual-note plans. It extracts the main ideas, divides them into frames, and writes English prompts for styles such as mind maps, handwritten notes, architecture diagrams, and comparison grids.
native-agent-swarms
Coordinate small teams of specialized Codex agents with bounded parallelism, explicit ownership, native messaging, and evidence-based synthesis. Use when two or more independent investigations, reviews, or implementation streams can run concurrently without sharing writable files, or when the user asks for swarm…
channel-truth-run
Automatic channel and contact synchronization.