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 GregoryHo/arcforge --skill speccinggit clone --depth 1 https://github.com/GregoryHo/arcforgeWrote 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/gregoryho/arcforge/speccing)<a href="https://agentmods.dev/skills/gregoryho/arcforge/speccing"><img src="https://agentmods.dev/badge/skills/gregoryho/arcforge/speccing/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/gregoryho/arcforge/speccing"><img src="https://agentmods.dev/badge/skills/gregoryho/arcforge/speccing.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 | $0.00061 | $0.02068 |
| Opus 5.5 | $0.00024 | $0.00827 |
| Sonnet 5.5 | $0.00012 | $0.00414 |
| Haiku 4.5 | $0.00006 | $0.00207 |
Grade A, and why
speccing 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 today.
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 — 174 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Speccing
Some repositories keep their product intent in files: a living spec per area, a versioned roadmap, an append-only decision log, and a backlog of wishes. Those files are the product's memory. This skill is how you work without corrupting it.
The law
THE SPEC AND THE CODE MERGE IN THE SAME PR
A RECORDED DECISION IS NEVER EDITED
A spec that lags the code is worse than no spec: the next reader trusts it and builds on a promise the code stopped keeping. A decision log you edit records what you currently believe rather than what you chose and why — which is the one thing it existed to preserve.
When this does not apply
- The repo keeps no product state and the user did not ask for any. Do not create it because it would be tidy — see step 1.
- The design is still open. Settle what to build with
/brainstorming; this skill records what was settled. - The change has no product footprint: a typo, a refactor with identical behavior, a dependency bump.
1. Bootstrap only when asked
Starting the ledger is a commitment the user makes. Offer once when the fit is obvious, then wait for a yes.
Once asked, four files: product/ROADMAP.md (the version table and the decision
log), product/BACKLOG.md (the wishes), product/specs/<slug>.md (one per
area), and a short guide saying how the other three are maintained. Shapes for
all four: references/templates.md.
2. Capture a wish
An idea nobody has committed to is one line in the backlog — a slug, a sentence, and an optional dependency note. No estimate, no spec, no roadmap row. Friction here has to stay near zero, or the good ideas go into a chat window instead.
3. Promote before building
Promotion is three edits — two, absent a wish — made before any code:
- If it started as a wish, replace its backlog line with a graduation tombstone
naming where it went (
references/conventions.md). Work the user asked for directly never had a line, and skips this edit alone. - Append a decision recording which version, and why now.
- Add the roadmap row and its spec — written from the template for a new area,
or this version's behaviors added to the area's existing living spec, never a
second file. Either way the spec's Decisions section carries that
decision's id and the one-line reason it is cited — the first entry on a new
spec, an addition on an existing one, never a swap. An existing spec's header
follows its highest linking row from this edit on, not from the build — the
compound form in
references/conventions.mdonce a lower row has shipped, plaindraftwhile none has. The new row takes← we are hereonly when every row above it has shipped; abuildingrow, or an earliernextrow, keeps it.
What ships with it
2 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.
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.
- today First seen · 174 lines · 61 tokens per session scan A e2784518293b
speccing is a skill published in the GitHub repository GregoryHo/arcforge (8 stars, last pushed 12d ago), licensed MIT. It adds 61 tokens to every session and 2,068 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 2026-11-07.
speccing compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| speccingGregoryHo/arcforge | Skill | 8 | 12d ago | MIT | A | 61 | Claude Code |
| imagegenopenai/codex | Skill | 133k | yesterday | Apache-2.0 | A | 113 | Codex |
| playwright-climicrosoft/playwright | Skill | 99k | yesterday | Apache-2.0 | A | 19 | Claude Code |
| local-ai-agentsmicrosoft/ai-agents-for-beginners | Skill | 79k | 4d ago | MIT | A | 200 | Claude Code, Codex |
| systematic-debuggingobra/superpowers | Skill | 313k | 3d ago | MIT | A | 21 | Claude Code |
| next-cache-components-adoptionvercel/next.js | Skill | 143k | yesterday | MIT | A | 95 | Claude Code |
| next-partial-prefetching-adoptionvercel/next.js | Skill | 143k | yesterday | MIT | A | 103 | Claude Code |