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 agentmods add skills/nick-railsback/skill-engine/inspect-ai-contextnpx skills add nick-railsback/skill-engine --skill inspect-ai-contextgit clone --depth 1 https://github.com/nick-railsback/skill-engineWrote 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/nick-railsback/skill-engine/inspect-ai-context)<a href="https://agentmods.dev/skills/nick-railsback/skill-engine/inspect-ai-context"><img src="https://agentmods.dev/badge/skills/nick-railsback/skill-engine/inspect-ai-context.svg" alt="Measured on agentmods" 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 | $0.00031 | $0.03237 |
| Opus 5 | $0.00015 | $0.01618 |
| Sonnet 5 | $0.00006 | $0.00647 |
| Haiku 4.5 | $0.00003 | $0.00324 |
Grade A, and why
inspect-ai-context 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 5d 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 — 109 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Context navigator (multi-source)
Overview
This navigator catalogs references for multiple sources the engine has been pointed at. Each source's references are catalogued in their own ## Catalog: <source-slug> section below; references live in references/<source-slug>-<topic>.md so the filename prefix discriminates sources at-a-glance.
The navigator's standing instructions stay small; references load only when relevant to the current question. The catalog itself is a TOC and is excluded from the standing-instructions budget, so additional sources and rows can be added without trimming prose.
When asked a question:
- Identify which source the question is about — see "How to search this navigator" below.
- Scan that source's Catalog section for the matching topic.
- Follow the link to read the reference file.
- If the question spans multiple sources, consult the Cross-source map.
Claims policy
Cite by default, and make load-bearing claims verifiable:
- Inline-cite every load-bearing claim with its SHA-pinned permalink — the
https://github.com/<owner>/<repo>/blob/<sha>/<path>#L<start>-L<end>link the reference gives for that fact (versions, defaults, signatures, deprecations, behavior a user could get wrong by guessing). Put the permalink inline, on the claim. Use a bare filename parenthetical (e.g.(<source-slug>-<topic>.md)) only when the reference genuinely provides no permalink. This inline permalink is what the grounded-citation eval (SELF-AUDIT Check 8) grades. - Don't cite orientational prose — "what is X?", "when did X launch?" — answer those from this navigator alone; opening a reference is itself a citation gesture.
- End with a one-line provenance footer, emitted italic, formatted
*References consulted: foo.md, bar.md. Grounded in {{LIBRARY}}@{{VERSION}} — [reference index]({{INDEX_URL}}).*The footer is a summary of what you read — not a substitute for the inline permalinks on the claims. Tokens are agent-substituted at answer time (they appear literally in the stampedSKILL.md); across multiple sources,{{LIBRARY}}resolves to the source slug(s) the answer drew from, or every registered slug when no reference was opened. - If no reference was opened, say so in the footer ("Answered from general knowledge — no {{LIBRARY}} references consulted") — never fake it.
What ships with it
23 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.
- references/inspect-aisi-org-uk-api-reference.md 10 KB
- references/inspect-aisi-org-uk-cli-reference.md 8.8 KB
- references/inspect-aisi-org-uk-docs-portal.md 5.1 KB
- references/ukgovernmentbeis-inspect-ai-agents.md 20 KB
- references/ukgovernmentbeis-inspect-ai-cli-and-config.md 12 KB
- references/ukgovernmentbeis-inspect-ai-datasets.md 8.4 KB
- references/ukgovernmentbeis-inspect-ai-eval-sets.md 17 KB
- references/ukgovernmentbeis-inspect-ai-extensions.md 17 KB
- references/ukgovernmentbeis-inspect-ai-logs-and-analysis.md 22 KB
- references/ukgovernmentbeis-inspect-ai-models-and-providers.md 20 KB
- references/ukgovernmentbeis-inspect-ai-overview.md 9.1 KB
- references/ukgovernmentbeis-inspect-ai-sandboxing.md 14 KB
- references/ukgovernmentbeis-inspect-ai-scorers.md 15 KB
- references/ukgovernmentbeis-inspect-ai-solvers.md 11 KB
- references/ukgovernmentbeis-inspect-ai-tasks.md 12 KB
- references/ukgovernmentbeis-inspect-ai-tools.md 15 KB
- research/.discover-cache.json 296 B
- research/.engine-stats.json 224 B
- research/.research-state.json 22 B
- research/review-state.json 97 B
- research/sessions/selfaudit-20260528T001740Z.json 247 B
- research/source-paths.json 907 B
- verify.sh 61 KB runs code
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.
- 5d ago First seen · 109 lines · 31 tokens per session scan A 3878e566208c
inspect-ai-context is a skill published in the GitHub repository nick-railsback/skill-engine (2 stars, last pushed today), licensed MIT. It adds 31 tokens to every session and 3,237 once invoked, about $0.0002 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
erpaval
Workflow for autonomous software development with the Kiro CLI (open Agent Skills standard). Classifiers route scope, complexity, directory state, and spec readiness before committing to Explore / Research / Plan / Act / Validate, then close the loop with a Compound step that persists lessons to .erpaval/solutions/…
product-discovery
File-first orchestrator for product discovery and spec frameworks. Frames the problem with a Discovery Lead, Product Analyst, or System Architect role. Spawns parallel framework agents (Double Diamond, How Might We, Jobs to Be Done, user stories with INVEST, EARS for AI-coder specs, Gherkin), synthesizes into a PRD or…
meta-prompt-optimizer
Audit and rewrite prompts against Claude Opus 4.7 instruction-following best practices. Produces a scored audit plus a rewritten prompt. Use when the user asks to audit, review, critique, rewrite, improve, optimize, tighten, score, fix, migrate, or draft a prompt — including system prompts, CLAUDE.md files, Claude…
product-strategy
File-first orchestrator for business and product strategy frameworks. Frames the challenge with a CPO / CSO / VP Design role, spawns parallel framework agents (Rumelt kernel + Crux, Wardley Maps, Minto Pyramid as thinking structure, Working Backwards PR-FAQ as discovery artifact), synthesizes outputs into one strategy…
agent-ux-patterns
Agent UX patterns and human-in-the-loop design: Twilio A2H protocol, Levels of Autonomy (L0-L5), inbox pattern, progressive trust, decision journal, gate reviews, autonomous work reports. Six-phase methodology from vague pain point to concrete data model via landscape research and multi-direction brainstorming. Use…
tech-stack-builder
Opinionated tech stack recommendations and settled library references. Two modes: (1) /build-stack runs parallel researcher agents to compare candidates and emit a stack report + ADRs for greenfield projects; (2) direct reference routing for coding with settled defaults. Use when the user asks to recommend a stack…