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 ConnorRX56/presentation-delivery-skills --skill human-intent-auditgit clone --depth 1 https://github.com/ConnorRX56/presentation-delivery-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/connorrx56/presentation-delivery-skills/human-intent-audit)<a href="https://agentmods.dev/skills/connorrx56/presentation-delivery-skills/human-intent-audit"><img src="https://agentmods.dev/badge/skills/connorrx56/presentation-delivery-skills/human-intent-audit/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/connorrx56/presentation-delivery-skills/human-intent-audit"><img src="https://agentmods.dev/badge/skills/connorrx56/presentation-delivery-skills/human-intent-audit.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 | $0.00077 | $0.00840 |
| Opus 5 | $0.00039 | $0.00420 |
| Sonnet 5 | $0.00015 | $0.00168 |
| Haiku 4.5 | $0.00008 | $0.00084 |
Grade A, and why
human-intent-audit 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 11d 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 — 92 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Human Intent Audit
Role
Own the translation from human goals to machine-verifiable evidence. Never make the user invent metrics, thresholds, traces, or implementation vocabulary unless a genuine product choice belongs to them.
[RULE:HUMAN-INTENT-001] A machine-green proxy cannot establish a human-goal PASS without a proved semantic link to the final user-observable outcome. WHY: A machine-green proxy cannot prove that the delivered outcome satisfies the user's natural-language requirement.
Input Contract
Collect the user's natural-language requirements, authoritative references, available artifacts, execution context, and any explicit non-goals. Preserve loaded terms such as “same”, “all”, “smooth”, “usable”, “safe”, “complete”, or “原封不动”; do not silently narrow them to whatever is easiest to measure.
Workflow
- Record each material human requirement and the interpreted intent.
- Trace source/reference -> producer -> consumer -> final visible or operational outcome. Mark every missing edge.
- Select final observables. Treat configured values, intermediate producers, state names, screenshots, and terminal values as proxies until equivalence is demonstrated.
- Derive predicates and tolerances from authoritative references, repeatability, perceptual or operational bounds, and risk. Never choose a tolerance merely because current output passes it.
- Run machine checks through the user's real path, then execute an owner-readable scenario without implementation jargon.
- Run a negative fixture reproducing the prior false-positive pattern.
- Audit both implementation-vs-predicate and predicate-vs-human-intent.
Perform the scenario before asking the owner to test. The owner is final authority, not the first QA worker or detector author.
Output Contract
For every requirement, record:
human_requirementandinterpreted_intent;final_observablesandmachine_predicates;threshold_basis;proxy_risksand unresolved semantic gaps;- machine
evidence; owner_acceptance_scenario;- status:
pass,partial,open,blocked, orfail.
What ships with it
5 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.
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.
- 11d ago First seen · 92 lines · 77 tokens per session scan A 873d032a9a1a
human-intent-audit is a skill published in the GitHub repository ConnorRX56/presentation-delivery-skills (11 stars, last pushed 16d ago), licensed MIT. It adds 77 tokens to every session and 840 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 2026-08-31.
Other skills, from other repositories
pptx-html-fidelity-audit
Audit a python-pptx export against its source HTML deck, identify layout/content drift (footer overflow, cropped content, missing italic/em, lost styling, off-rhythm spacing), and re-export with strict footer-rail + cursor-flow layout discipline. Use this skill whenever the user has a .pptx that was generated from an…
build-test
Run the project's build / typecheck / lint / test commands and emit the build.passing + tests.passing signals devloop convergence reads.
playwright-cli
Automates browser interactions for testing and validating your own web applications using playwright-cli. Use when you need terminal-first browser control for navigation, form filling, screenshots, tracing, bound browser sessions, debugging, or generating Playwright test code. Only use against applications you own or…
design-ship
One-shot pipeline turning a claude.ai/design link into a pull request: scaffold via /ork:design-import, stories and specs via /ork:cover, browser verification via /ork:expect, then open the PR. Use when a design link should come back as a PR with no intermediate steps; if all you need is the components written to…
auto-test-code
A structured process for critically reviewing and testing software code. It records review findings, test plans, commands, results, and supporting files in a project workspace.
auto-test-skill
A structured process for critically testing and improving a coding skill. It records test plans, findings, fixes, and quality checks in files that can be reviewed later.