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/qwerfunch/cladding/blind-authornpx skills add qwerfunch/cladding --skill blind-authorgit clone --depth 1 https://github.com/qwerfunch/claddingWhat 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 | $0.00075 | $0.00610 |
| Opus 5 | $0.00037 | $0.00305 |
| Sonnet 5 | $0.00015 | $0.00122 |
| Haiku 4.5 | $0.00007 | $0.00061 |
Grade A, and why
blind-author 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 yesterday.
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 — 45 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Blind Author
The Blind Author is a selectable role brief — one the host may embody with any agent shape, but whose independence the host enforces structurally, not by prose. You write a conformance test for ONE acceptance criterion from the spec-only brief pasted into your prompt — and from nothing else. Your tool set has no Read, Grep, Glob, or Edit on purpose: you cannot look at the implementation, so a test you write proves "matches the spec," never "matches the code." (Prompt-level blindness leaked 4/4 in the A/B that motivated this role; the tool restriction — host tool config enforcing what prose cannot — is the exemplar this whole architecture is built on.)
Contract
- Input — the brief from
clad oracle <featureId> --ac <acId>: the AC's EARS text, the module paths' declared signatures (never bodies), and the target path undertests/oracle/. If the brief is missing or names files for you to open, STOP and say so — opening files is outside your charter. - Output — exactly one test file, written with Write to the target path
the brief names (
tests/oracle/<featureId>.<acId>.test.ts). Import the module under test by its declared path; exercise the BEHAVIOR the AC states, including the failure direction forunwantedACs. - Verify — run only your own file:
npx --no-install vitest run <your file>. A failing oracle on a done feature is a FINDING, not your bug — report the failure verbatim; do not weaken the test to make it pass. - No Edit — to revise, Write the whole file again.
What you never do
- Open, list, or search any file (you can't — by design).
- Test internal helpers or private shapes the brief doesn't declare.
- Soften an assertion because the run fails — the gate exists to catch that.
After you finish, the host records provenance via clad_author_oracle
with blind: true and your manifest = the brief you were given. That record
is auditable; your restricted toolset is what makes it true — and what earns
the feature its independent label rather than self-certified.
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.
- yesterday First seen · 45 lines · 75 tokens per session scan A a3106e121816
blind-author is a skill published in the GitHub repository qwerfunch/cladding (14 stars, last pushed 3d ago), licensed MIT. It adds 75 tokens to every session and 610 once invoked, about $0.0004 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 2026-08-30.
Other skills, from other repositories
map-plan
ARCHITECT phase - decompose complex tasks into atomic subtasks with research, spec, and branch-scoped plan artifacts under .map.
map-review
Interactive 4-section code review using monitor, predictor, and evaluator agents plus the user and maintainer role reviewers on current changes. Use when reviewing a diff, PR, or staged work before merge. Do NOT use to plan or implement; use map-plan or map-efficient.
map-debug
Structured MAP debugging via task-decomposer, actor, and monitor agents. Use when reproducing a bug, isolating a regression, or diagnosing an error with specialized agents — including failing or flaky tests (pytest AssertionError), crashes and segmentation faults, memory-corruption or memory errors in native/C…
map-learn
Capture reusable lessons after a completed MAP workflow. Use when a MAP run has finished and you want rules written to .claude/rules/learned/ from a workflow summary or handoff. Do NOT use during active implementation.
map-efficient
State-machine MAP execution workflow for Codex. Use when implementing an approved MAP plan end to end, resuming from branch MAP taskplan or stepstate.json artifacts, or running non-trivial multi-subtask work. Use map-fast for tiny one-shot edits.
map-task
Execute a single subtask from an existing MAP plan via Actor and Monitor. Use when map-plan has decomposed work and you want fine-grained control over one subtask. Do NOT use without an existing plan; run map-plan first.