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 arcasilesgroup/ai-engineering --skill spec-extractgit clone --depth 1 https://github.com/arcasilesgroup/ai-engineeringWrote 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/arcasilesgroup/ai-engineering/spec-extract)<a href="https://agentmods.dev/skills/arcasilesgroup/ai-engineering/spec-extract"><img src="https://agentmods.dev/badge/skills/arcasilesgroup/ai-engineering/spec-extract/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/arcasilesgroup/ai-engineering/spec-extract"><img src="https://agentmods.dev/badge/skills/arcasilesgroup/ai-engineering/spec-extract.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 | $0.00148 | $0.02235 |
| Opus 5.5 | $0.00059 | $0.00894 |
| Sonnet 5.5 | $0.00030 | $0.00447 |
| Haiku 4.5 | $0.00015 | $0.00224 |
Grade A, and why
spec-extract 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 1mo 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 — 185 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Extract a spec from what was actually agreed
design-check halts without a spec. verify-feature falls back to reconstructing one
from conversation memory, which is the weakest contract it supports. Both problems are the
same problem: the agreement exists, but only in a chat log that nobody will re-read and
that the next session cannot see.
Your job is to move it out of the conversation and into a file, without inventing anything. The single failure mode of this skill is a confident spec containing requirements nobody ever asked for — that is worse than no spec, because every downstream verification then measures against fiction and reports it as fact.
1. Do not read the implementation first
If code already exists for this work, do not read it before you have drafted the
requirement list. You will reconstruct requirements that the code happens to satisfy,
and produce a spec that is guaranteed to pass its own check. This is the same ordering
rule as verify-feature §1 and it matters more here, because the artifact persists.
There is one legitimate exception, and it must be labeled. When the goal is explicitly to document something already built — a legacy subsystem, an inherited codebase — you are writing a descriptive spec, not a prescriptive one. Say so in the header, because a descriptive spec cannot detect drift that already happened; it canonizes it.
2. Gather sources in order of authority
- The user's own messages in this session — the most common source, and the one that disappears when the session ends. Read the whole session, not the opening ask.
- A ticket or issue —
gh issue view,glab issue view, a linked doc, a pasted thread. Quote it rather than paraphrase where the wording is load-bearing. - Existing partial docs — a half-written
design.md, an ADR, a README section, a TODO list. These are prior agreements; do not silently overwrite them. - Project invariants that constrain this work —
CLAUDE.md,AGENTS.md,CONTRIBUTING.md, the schema, the auth convention. These are requirements even though nobody restated them for this feature.
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.
- 1mo ago First seen · 185 lines · 148 tokens per session scan A fb2a8ef42e96
spec-extract is a skill published in the GitHub repository arcasilesgroup/ai-engineering (61 stars, last pushed 2d ago), licensed Apache-2.0. It adds 148 tokens to every session and 2,235 once invoked, about $0.0006 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 2026-09-12.
Other skills, from other repositories
spec-kitty-charter-doctrine
Run charter interview, generation, context, and sync workflows for project governance in Spec Kitty 3.x. Access doctrine artifacts programmatically via DoctrineService. Resolve agent profiles. Load action-scoped governance context iteratively, not all at once. Triggers: "interview for charter", "generate charter"…
322-frameworks-spring-boot-testing-integration-tests
Use when you need to write or improve integration tests — including Testcontainers with @ServiceConnection, @DataJdbcTest persistence slices, TestRestTemplate or MockMvcTester for HTTP, data isolation, and container lifecycle management for Spring Boot 4.0.x. This should trigger for requests such as Review Java code…
setup-husky-dotnet
Use when configuring Git hooks in .NET projects before team commits occur, to enforce commit message standards and code formatting automatically.
314-frameworks-spring-kafka
Use when you need to design or implement Kafka messaging in Spring Boot — including topic design, producer/consumer implementation, JSON serialization with Boot factory customizers, Testcontainers @ServiceConnection integration tests, retries and dead-letter topics, idempotency, and error handling. This should trigger…
audit
Independent, on-demand AL codebase auditor for Business Central. Judges the code against BCQuality (citable knowledge) plus native checks for what BCQuality does not reach. Read-only; advisory verdict. Default scope: objects changed vs main; full codebase on request.
review
WCAG 2.2 AA curated accessibility review — distinguishes automatically detectable, browser-assisted, and manual/human-judgment findings with evidence citations and SC mapping. Composes with design-assessment/design-improvement/frontend-design-review.