Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/leopu00/job-hunter-teamnpx agentmods add skills/leopu00/job-hunter-team/cv-disk-auditWrote 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/leopu00/job-hunter-team/cv-disk-audit)<a href="https://agentmods.dev/skills/leopu00/job-hunter-team/cv-disk-audit"><img src="https://agentmods.dev/badge/skills/leopu00/job-hunter-team/cv-disk-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/leopu00/job-hunter-team/cv-disk-audit"><img src="https://agentmods.dev/badge/skills/leopu00/job-hunter-team/cv-disk-audit.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00091 | $0.01234 |
| Opus 5 | $0.00046 | $0.00617 |
| Sonnet 5 | $0.00018 | $0.00247 |
| Haiku 4.5 | $0.00009 | $0.00123 |
Grade A, and why
cv-disk-audit scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 91 lines — stays where its author put it; the contents beside it link to each section on GitHub.
cv-disk-audit — riconciliazione disk↔DB sui CV
Bug #26 ha mostrato il pattern: lo Scrittore genera il PDF, viene
killato (EMERGENZA freeze 2026-05-17 04:43) prima dell'UPDATE DB. Il
file resta su /jht_user/cv/, ma applications.cv_pdf_path resta NULL.
Sisal 7.5/10 (top PASS della finestra) era diventato "CV da scrivere"
sulla dashboard utente — invisibile.
Il fix preventivo (atomic write in cv-structure skill) impedisce
nuovi orfani. Questa audit ricuce quelli già esistenti e cattura ogni
nuova divergenza che dovesse comparire (es. utente sposta a mano un
PDF, watchdog uccide il Writer durante il rename).
Quando lanciarla
Trigger Dottore (end-of-round, fuori budget critico):
- Sempre nel primo giro dopo un EMERGENZA / kill di uno Scrittore.
- Altrimenti ~ogni 4 giri Dottore (≈2h, dato il giro 30 min).
Il Dottore esegue questa skill DOPO liveness-check e PRIMA di
cache-prune — l'audit è informativo, non distruttivo.
Procedura
# 1. Snapshot disk
DISK_PDFS=$(find /jht_user/cv -maxdepth 1 -type f -name '*.pdf' 2>/dev/null | sort)
# 2. Snapshot DB (cv_pdf_path != NULL)
DB_PDFS=$(python3 /app/shared/skills/db_query.py cv-pdf-paths 2>/dev/null | sort)
# 3. Diff
ORFANI=$(comm -23 <(echo "$DISK_PDFS") <(echo "$DB_PDFS")) # disk ma non DB
GHOST=$(comm -13 <(echo "$DISK_PDFS") <(echo "$DB_PDFS")) # DB ma non disk
# 4. Report al Capitano (deterministico, niente LLM)
if [ -n "$ORFANI$GHOST" ]; then
msg="[@dottore -> @capitano] [REPORT] CV audit mismatch — "
msg="${msg}orfani=$(echo "$ORFANI" | grep -c .) "
msg="${msg}ghost=$(echo "$GHOST" | grep -c .)"
jht-tmux-send CAPITANO "$msg"
# Log dettagli
ts=$(date -u +%Y-%m-%dT%H:%M:%SZ)
echo "{\"ts\":\"$ts\",\"orfani\":$(echo "$ORFANI" | jq -R . | jq -s .),\"ghost\":$(echo "$GHOST" | jq -R . | jq -s .)}" \
>> /jht_home/logs/cv-disk-audit.jsonl
fi
db_query.py cv-pdf-paths (da implementare): scrive 1 path per riga di
tutte le applications con cv_pdf_path IS NOT NULL. Una linea
script-friendly per il comm.
What ships with it
6 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 · 91 lines · 91 tokens per session scan A e8de13e859a7
cv-disk-audit is a skill published in the GitHub repository leopu00/job-hunter-team (49 stars, last pushed yesterday), licensed MIT. It adds 91 tokens to every session and 1,234 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 2026-08-30.
Other skills, from other repositories
database
A database design and migration guide written in Korean. It covers table structure, naming conventions, common columns, indexes, foreign keys, and versioned migration files.
db-migrate
Creates, checks the status of, and validates database migrations by auto-detecting the project's migration framework (Alembic, Prisma, Knex, Django, Rails/ActiveRecord, Flyway, Atlas, TypeORM, Sequelize, SQLx, golang-migrate, Liquibase) from marker files, then driving that framework's own CLI. Use when the user wants…
airtable
Airtable base management — records, views, and formulas via REST API.
landing-performance
Fix performance issues in apps/landing for leaderboards, battles, rankings, vote/reaction summaries, and high-cardinality result pages.
storage-steer
Prioritize hardening persistent storage mechanisms in scenarios/{{TARGET}}/ to ensure reliable, isolated, and professionally structured data persistence. This skill steers toward storage that is correctly configured, properly abstracted, and follows Vrooli resource conventions.
sql-database
Safe schema-aware SQL querying against CSV/Parquet/JSON files, SQLite databases, and live Postgres. Use when querying structured data, profiling a schema, or building data pipelines.