Borrowing it
Nothing to install: this file belongs to AHepi/DeepReason. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/AHepi/DeepReason/main/.claude/skills/dr-audit-goal-trace/SKILL.mdgit clone --depth 1 https://github.com/AHepi/DeepReasonWrote 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/ahepi/deepreason/dr-audit-goal-trace)<a href="https://agentmods.dev/skills/ahepi/deepreason/dr-audit-goal-trace"><img src="https://agentmods.dev/badge/skills/ahepi/deepreason/dr-audit-goal-trace.svg" alt="Measured on agentmods" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00055 | $0.00845 |
| Opus 5 | $0.00028 | $0.00423 |
| Sonnet 5 | $0.00011 | $0.00169 |
| Haiku 4.5 | $0.00006 | $0.00085 |
Grade A, and why
dr-audit-goal-trace 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 8d 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 — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Audit: operator goals vs enforcement
Entry: LEDGER.md exists, goal-trace.md missing. Exit:
goal-trace.md written, one row per law, proofs in proof/.
The operator's goals live in two ledgers: CLAUDE.md §"Operator design laws" (standing, verbatim-quoted) and each tranche's REQUEST.md (per change). This worker traces the STANDING laws; a per-tranche trace is dr-validate-change's job, not this one.
A law counts as ENFORCED only when a mechanism would visibly fail if the law were violated — a test, a typed refusal/notice, a gate. Prose restating the law enforces nothing.
Operations
- Law census: extract every bolded law heading from CLAUDE.md
§"Operator design laws" into
proof/goal-laws.txt, one row each, with its date and verbatim kernel. - For each law, three scans, outputs saved to
proof/goal-<n>.txt: a. Mechanism scan:rg -l -i '<law key terms>' src/deepreason/b. Test scan:rg -l -i '<law key terms>' tests/c. Notice/refusal scan:rg -n '<law's typed string, if any>' src/ - Verdict per law, exactly one of:
enforced— row names the mechanism file AND the test that goes red on violation (both cited from the scans).partially-enforced— mechanism exists, no test pins it; the row names which half is missing.unenforced— scans empty; the law lives only in prose. The pasted empty scans are the proof (G2).process-law— governs agent/operator behavior, not code (e.g. a working-style rule); code enforcement is not expected. Sayingprocess-lawrequires one sentence naming WHO enforces it instead (the workflow file that carries it).
- For every
unenforcedandpartially-enforcedrow: PARK a prompt proposing the smallest mechanism that would make violation visible (a regression test, a typed notice), route dr-change-orchestrator. Proposing is this worker's ceiling — choosing is the operator's. - Write
goal-trace.md: the table, a count line, and one closing list: laws added since the last audit (diffproof/goal-laws.txtagainst the previous audit tranche's copy, if one exists).
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.
- 8d ago First seen · 68 lines · 55 tokens per session scan A 3ba1ba4f63ed
dr-audit-goal-trace is a skill published in the GitHub repository AHepi/DeepReason (142 stars, last pushed yesterday), licensed MIT. It adds 55 tokens to every session and 845 once invoked, about $0.0003 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
research-engineer
An uncompromising Academic Research Engineer. Operates with absolute scientific rigor, objective criticism, and zero flair. Focuses on theoretical correctness, formal verification, and optimal implementation across any required technology.
tika-eval-compare
Compare extracts from two Tika builds over a corpus to detect regressions in content, encoding, exceptions, and embedded-document handling. Use for "compare before/after extracts", "eval this change against the corpus".
neuron-evaluation-engineer
Create and run AI evaluations with datasets, assertions, and output drivers in Neuron AI. Use this skill whenever the user mentions evaluation, testing AI systems, creating evaluators, dataset-driven testing, assertion-based validation, or wants to measure AI system performance. Also trigger for tasks involving…
jetson-validate-image
Use after jetson-flash-image to run static BSP checks, on-target smoke/regression tests on a flashed DUT, or both. Not for build or flash steps. Triggers: validate bsp, on-target validation.
atmos-validation
Validate Atmos projects, components, arbitrary JSON Schema inputs, EditorConfig, and GitHub Actions; use affected-file selection and native CI annotations.
skill-benchmark
Benchmark AI skill effectiveness by measuring implementation quality against legacy constraints.