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 GreatMark/fable-field-guide-skills --skill blindspot-passgit clone --depth 1 https://github.com/GreatMark/fable-field-guide-skillsWrote 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/greatmark/fable-field-guide-skills/blindspot-pass)<a href="https://agentmods.dev/skills/greatmark/fable-field-guide-skills/blindspot-pass"><img src="https://agentmods.dev/badge/skills/greatmark/fable-field-guide-skills/blindspot-pass/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/greatmark/fable-field-guide-skills/blindspot-pass"><img src="https://agentmods.dev/badge/skills/greatmark/fable-field-guide-skills/blindspot-pass.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 | $0.00146 | $0.00977 |
| Opus 5 | $0.00073 | $0.00489 |
| Sonnet 5 | $0.00029 | $0.00195 |
| Haiku 4.5 | $0.00015 | $0.00098 |
Grade A, and why
blindspot-pass 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 10d 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 — 73 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Blindspot Pass
Help the user discover their unknown unknowns before work starts. This is a read-only pass: do not modify any files.
Workflow
- Anchor on the user's starting point. If not already clear from the conversation, ask once — a single short message, at most three sub-questions — about their experience with this problem and codebase, what they have already decided, and what they plan to do next. If asking is not possible (e.g. an unattended/CI run), the user deflects ("just go"), or the reply does not answer the questions, state your assumptions in one line — folding in whatever they did say — and proceed.
- Explore the territory. Start from the files the task will touch — or, when the user is adopting an area rather than doing one task, its entry points (public API, main module, README) — and expand outward: direct dependencies/dependents, tests, docs, then git history of the touched files (not the whole repo). If history is unavailable (tarball, shallow clone, fresh repo) or the area is not code, use its changelog, release notes, or prior internal write-ups instead. For non-code areas (a SaaS tool, a business decision), expand the same way through official docs and pricing, adoption and migration post-mortems from other teams, the vendor's changelog and viability, and compliance constraints. In large codebases, cap this pass at roughly 15-25 file reads/searches: follow the 2-3 leads most likely to change the user's approach; if a lead stalls (slow search, missing tooling), drop it after one retry; when you hit the cap, stop and report, listing unexplored areas explicitly as unswept corners. If the domain is unfamiliar to you or fast-moving, search the web. Look specifically for things the user did not mention in their description of the task.
- Report blindspots directly in the chat response, in the language of the
conversation (do not create files unless the user asks for a document),
organized as:
- Questions you didn't know to ask — decisions hidden in this area (existing conventions, invariants, feature flags, edge cases; for non-code: pricing-tier traps, data-export paths, compliance).
- What "good" looks like — quality bars, reference implementations, or prior art the user should calibrate against.
- History and potholes — past attempts, known pitfalls, TODO/FIXME landmines, coupling that will bite.
- Domain concepts — if the domain is new to the user, teach the 3-5 concepts they need to evaluate results (not an encyclopedia).
- End with "how to prompt me better": 3-6 concrete things the user should specify in their next prompt now that these blindspots are visible. If "what good looks like" surfaced a concrete reference worth copying, suggest reference-hunt. If the remaining ambiguities would change the approach, can only be decided by the user, the session is interactive, and the user has not declined to answer questions, suggest running interview-me next; otherwise end after the prompt advice. If the area was simple (see Rules), shrink or skip this section.
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.
- 10d ago First seen · 73 lines · 146 tokens per session scan A df762396dcfe
blindspot-pass is a skill published in the GitHub repository GreatMark/fable-field-guide-skills (9 stars, last pushed 1mo ago), licensed MIT. It adds 146 tokens to every session and 977 once invoked, about $0.0007 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
idea-refine
Refines raw ideas into sharp, actionable concepts through structured divergent and convergent thinking. Use when an idea is still vague, when you need to stress-test assumptions before committing to a plan, or when you want to expand options before converging on one. Triggers on "ideate", "refine this idea", or…
chinese-git-workflow
A reference for configuring Git with Chinese code-hosting services such as Gitee, Coding.net, GitLab China, and CNB, including SSH, HTTPS, credentials, CI, and repository mirroring.
chinese-documentation
A Chinese technical-documentation style guide covering spacing, punctuation, numbers, terminology, and links when Chinese and English appear together.
skillshare-implement-feature
Implement a feature from a spec file or description using TDD workflow. Use this skill whenever the user asks to: add a new CLI command, implement a feature from a spec, build new functionality, add a flag, create a new internal package, or write Go code for skillshare. This skill enforces test-first development…
setup-matt-pocock-skills
A setup skill that configures engineering skills for a repository, including its issue tracker, labels, and documentation layout. A repository is the project folder managed by version control.
create-slack-app
Use when a developer wants to create, scaffold, or bootstrap a new Slack app or agent from scratch with the Slack CLI. Covers prerequisites, sandbox setup, authentication, and creating + running a project from a Bolt (JS or Python) template locally. Trigger on "create a Slack app", "new Bolt app", "start a Slack…