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 somnus0x/agt-skill-pack --skill wrapgit clone --depth 1 https://github.com/somnus0x/agt-skill-packWrote 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/somnus0x/agt-skill-pack/wrap)<a href="https://agentmods.dev/skills/somnus0x/agt-skill-pack/wrap"><img src="https://agentmods.dev/badge/skills/somnus0x/agt-skill-pack/wrap/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/somnus0x/agt-skill-pack/wrap"><img src="https://agentmods.dev/badge/skills/somnus0x/agt-skill-pack/wrap.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 | $0.00095 | $0.01329 |
| Opus 5.5 | $0.00038 | $0.00532 |
| Sonnet 5.5 | $0.00019 | $0.00266 |
| Haiku 4.5 | $0.00010 | $0.00133 |
Grade B, and why
wrap scanned grade B with 2 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 17d 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
Then merge into `~/.claude/settings.json`: Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -sL $BASE/session-handoff.py -o ~/.claude/hooks/session-handoff.py How it starts
The opening of the file, as written. The whole thing — 109 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Wrap — handoff, then /clear
Why: a long main loop re-reads its whole context every call, so a short reply typed into a huge chat after an idle gap costs the same as a long one. One cold resume at 800K context runs $5–10 at list price. A ~50-line brief carries what matters at a fraction of that.
You cannot run /clear yourself — your job is to make clearing lossless, then tell the user to
type it.
Steps
- Path:
python3 ~/.claude/hooks/session-handoff.py --path→ the one handoff file for this cwd. Overwrite it; never write elsewhere (never into the repo). - Gather receipts, not memory (one batch of read-only calls):
git status --short,git branch --show-current,git log --oneline -5, branches ahead of origin — skip if not a git repo. Re-read any file you are about to cite. - Write the brief — terse, ≤ 60 lines, only what the next session can't cheaply re-derive from the repo:
# Wrap — <cwd basename> — <YYYY-MM-DD HH:MM>
**Next:** <the single next action; use $ARGUMENTS if given>
## State
- Doing: <task in 1-2 lines> · Branch: <b> · Uncommitted: <files or "clean">
- Done this session: <bullets, each with file:line or commit sha>
## Decisions (locked — don't relitigate)
- <decision> — because <why>
## Open
- <question / blocker / thing the user still has to decide>
## Gotchas
- <non-obvious things that cost time this session: wrong turns, corrected facts, tool quirks>
## Resume
- Files to read first: <paths>
- Commands: <exact commands to re-run / verify>
Rules: corrected facts go in Gotchas (e.g. "the staging DB is NOT a copy of prod"). No secrets, no key values. Don't paste code — point to file:line. Don't restate CLAUDE.md. 4. Surface stranded work: if this session created/modified files that are uncommitted, or a branch is ahead of origin, say so in one line — don't commit unless asked. 5. Hand over — reply with exactly:
- the handoff path and line count,
Type /clear — the next session loads this automatically.(also loads on a freshclaudestart in this directory within 24h),- if the next task is a taste/design call, suggest
/modelafter the clear; otherwise stay on the default.
What ships with it
3 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.
- 17d ago First seen · 109 lines · 95 tokens per session scan B dc67057c57c1
wrap is a skill published in the GitHub repository somnus0x/agt-skill-pack (74 stars, last pushed 18d ago), licensed Apache-2.0. It adds 95 tokens to every session and 1,329 once invoked, about $0.0004 per session on Opus 5.5. A static security scan graded it B with 2 findings (reads agent configuration directories, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-12.
Other skills, from other repositories
Migrate
Intakes external content (notes, CLAUDE.md, Cursor rules, Obsidian/Notion/Apple Notes exports), classifies chunks against the LifeOS taxonomy, and commits with provenance after confidence-tiered approval. USE WHEN /migrate, migrate content, import from other LifeOS, bring in old notes, import Cursor rules, import…
Cortex
LifeOS memory, prior-session recall, and learning capture. USE WHEN capture this learning, document this learning, what did we learn recently, recent learnings, cortex, knowledge, knowledge base, search knowledge, what do we know about, archive, harvest, knowledge status, develop note, add to knowledge, add a people…
Profile
View, edit, or clear the LifeOS User Model (PRINCIPALIDENTITY.md) — persistent identity context loaded at session startup. Three intents: View, Edit (identity, goals, constraints, preferences), Clear (reset). Read by Algorithm OBSERVE and skills for personalization. USE WHEN profile, show profile, view profile, who am…
wiki-dashboard
Create dynamic, queryable Obsidian dashboard views using Bases or Dataview. Use for auto-updating tables, indexes, or structured views of vault content; not for ordinary question answering.
relevance-coarse-filter
Cheap, high-recall first-pass filter that removes obvious junk from a detector candidate pool before expensive story-origin research and PR judgment. Decides keep, monitoronly, or reject — never ranks, writes angles, verifies dates, or decides whether to pitch.
session-compact
AI DevKit · Compact a historical AI coding session with Jev when context is long, work is handed off or resumed, or durable continuation facts and memory candidates need extraction.