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 donald-ada/workinggenius --skill blindspotgit clone --depth 1 https://github.com/donald-ada/workinggeniusWrote 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/donald-ada/workinggenius/blindspot)<a href="https://agentmods.dev/skills/donald-ada/workinggenius/blindspot"><img src="https://agentmods.dev/badge/skills/donald-ada/workinggenius/blindspot.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 | $0.00090 | $0.00511 |
| Opus 5 | $0.00045 | $0.00255 |
| Sonnet 5 | $0.00018 | $0.00102 |
| Haiku 4.5 | $0.00009 | $0.00051 |
Grade A, and why
blindspot 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 6d 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
Blindspot
The map — your prompt, your plan, your glossary — is not the territory: the codebase, its history, its real constraints. The gap between them is the unknowns, and an unknown left unfound doesn't stay unfound: the diff review finds it, or production does.
The concept: go look, at the three moments the gap is widest.
- Before unfamiliar work — walk the territory, read-only, as a fresh frontier-tier subagent (hunting unknown unknowns is judgment, not reading — measured twice; and the fresh context means the main session consumes the report instead of re-walking the files). The code, its tests, its history: where it bit last time is the best predictor of where it bites next. Report whatever matters — questions nobody thought to ask, what "good" looks like here, potholes, a sharper ask, and anything that fits no bucket at all; no taxonomy limits what counts as a finding. Every finding carries its evidence, and potholes are found, never manufactured. Check your own recommendations against your own findings before offering them.
- At a choice the user can't evaluate — "whichever you think" is a judgment gap, not agreement. Teach just enough of the actual difference to hold an opinion, then re-present the choice; if they still can't call it, an honest
assumed:beats a hollow confirmation. - Before acceptance — nothing verifies the user's map of what got built. Summarize what changed at the behavior level, then quiz the consequences they'll live with. A wrong answer is a finding about the explanation, not the user: repair the map, ask again — so they accept knowing what they accepted.
Standalone or inside a larger flow, the moves are the same. Findings route to whoever is shaping the work: questions into the conversation defining it, terms into the project's glossary where it keeps one, hazards to whoever judges the design.
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.
- 6d ago First seen · 18 lines · 90 tokens per session scan A 1e843772da53
blindspot is a skill published in the GitHub repository donald-ada/workinggenius (6 stars, last pushed today), licensed MIT. It adds 90 tokens to every session and 511 once invoked, about $0.0005 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
self-improve
Self-improve this skills repo — surface and (safely) apply quality improvements across four parallel read-only lenses (Clarity, Leanness, Coverage, Ergonomics). Sibling of /self-audit — audit fixes rule violations, improve raises quality where no rule is broken. Local meta-skill, independent of the /task: pipeline.
self-audit
Self-audit this skills repo against CLAUDE.md invariants, the artifact contract, and README/CLAUDE.md/docs sync via three parallel read-only subagents. Local meta-skill — independent of the /task: pipeline.
grill
Interrogate a plan or decision one question at a time before capture, keeping a decision-plus-rationale ledger, then route to the right capture skill.
dw-doctor
Read-only environment diagnostic for a solo-lane repo: whether the tools the hooks and skills assume are installed and whether the repo's guardrails will really fire, each gap reported with a copy-paste fix.
dw-handoff
Compact the live session into .ai/work/ /HANDOFF.md — how far into the current task you are, what is applied but uncommitted, and which dead ends are already ruled out, so the next context window resumes mid-task. Explicit-invoke only.
dw-next
The solo lane's build step and its resume point in one skill: report where the active .ai/work/ /CHANGE.md stands — read from disk, so it survives a /clear — then build every remaining unticked task, one commit each.