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 commands/agoldian/claudenv/deepergit clone --depth 1 https://github.com/AGoldian/claudenvWhat 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 | $0.00024 | $0.00531 |
| Opus 5 | $0.00012 | $0.00266 |
| Sonnet 5 | $0.00005 | $0.00106 |
| Haiku 4.5 | $0.00002 | $0.00053 |
Grade A, and why
deeper 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 2d 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
/deeper — Подробное объяснение последнего решения
Развернуть последнее (или указанное в $ARGUMENTS) принятое vibe-decision в полный формат.
Step 1 — Find the target decision
If $ARGUMENTS is empty:
- Run
claudenv decisions list --limit 1(Bash) — get the most recent decision id - Run
claudenv decisions show <id>to load its frontmatter
If $ARGUMENTS is a topic slug or id:
- Run
claudenv decisions show $ARGUMENTSto load it - If not found — fall back to
claudenv decisions search "$ARGUMENTS"and pick the top match
If nothing matches:
Не нашёл недавнего решения. Сначала сделай нетривиальный выбор — vibe-decisions залогирует его. Потом вызови /deeper.
Stop here.
Step 2 — Expand into the deep dive format
Follow ~/.claude/skills/vibe-decisions/deep-dive-template.md exactly:
- Концептуально (3-5 lines) — what it is, what problem it solves, name 1-2 alternatives
- Как работает (5-10 lines, optional code sketch) — simplified model + key invariants
- 2-3 варианта реализации — for each: code sketch + when-suits / when-doesn't
- Канон — 2-4 references, first from
~/.claudenv/memories/canon/index.yamlby topic; if no match, WebSearch and proposeclaudenv canon addafterward
Step 3 — Update the decision file
Edit the original decision file to set:
deep_dive_done: yes
sources_consulted: [<urls and canon ids you cited>]
Preserve the __VIBE_DECISION__ marker on the last line.
Style
- User's language
- Concrete to the actual task, not generic textbook
- If a sketch is shorter than prose — use the sketch
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.
- 2d ago First seen · 56 lines · 24 tokens per session scan A adec1e85d291
deeper is a command published in the GitHub repository AGoldian/claudenv (10 stars, last pushed 2mo ago), licensed MIT. It adds 24 tokens to every session and 531 once invoked, about $0.0001 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.