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 skills/drn/dots/handoffnpx skills add drn/dots --skill handoffgit clone --depth 1 https://github.com/drn/dotsWhat 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 | $0.00031 | $0.03793 |
| Opus 5 | $0.00015 | $0.01896 |
| Sonnet 5 | $0.00006 | $0.00759 |
| Haiku 4.5 | $0.00003 | $0.00379 |
Grade A, and why
handoff 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.
How it starts
The opening of the file, as written. The whole thing — 207 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Context Handoff
Generate a structured prompt capturing the current conversation context so it can be pasted into another agent thread.
Arguments
$ARGUMENTS- Optional. Free-form instructions about what to emphasize or who the target is. May also specify the target Argus project the follow-up task is created in, two ways:project=<name>— explicit override, recognized only as a standalone whitespace-delimited token (so the word "project" in a sentence is not parsed as a directive). Used verbatim and takes precedence over any cue-anchored target. It is intentionally not validated against the projects list: the explicitproject=syntax signals deliberate intent, so the user's literal value is honored.- A cue-anchored target: the first whitespace-delimited token after
to,for,hand off to, orhandoff to— for example/handoff to keystone this taskor/handoff this work to keystone. Compare that token (lowercased, with surrounding punctuation stripped) against the Argus projects listed in the Context block. A token is treated as the target only when it matches a known project; the routing phrase is then ignored when assembling the emphasis content. If several cue phrases appear, the first one whose token matches a known project wins. A cue-anchored token that matches no known project is never silently used — the skill asks (see step 8). Becauseto/forare common words, this validation is what prevents ordinary prose (e.g. "focus on the api changes", "remember to test") from being mistaken for routing: prose only routes if a word right after a cue happens to name a real project, and even then the handoff body is shown for review before any task is created. - If no project is specified either way, derive from the worktree path: the immediate subdirectory after
~/.argus/worktrees/(e.g. CWD/Users/me/.argus/worktrees/dots/foo→ projectdots). If that does not resolve, ask the user before creating the task. no-task— standalone token that skips Argus task creation entirely (KB save still runs). When present it short-circuits step 8 before any project resolution, so a cue-anchored orproject=target is moot and the skill does not ask about an unknown project.
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 · 207 lines · 31 tokens per session scan A e3c637a5fade
handoff is a skill published in the GitHub repository drn/dots (23 stars, last pushed 4d ago), licensed MIT. It adds 31 tokens to every session and 3,793 once invoked, about $0.0002 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
mcporter
List, auth, and call MCP servers/tools from the terminal.
agent-messaging
Send and receive cryptographically signed messages between AI agents using the Agent Messaging Protocol (AMP). Use when the user asks to "send a message to an agent", "check agent inbox", "message another agent", "reply to a message", "notify an agent", or any inter-agent communication task.
📝 任务完成后归档
重要提醒: 每次完成复杂调试或开发任务后，主动执行此流程！ 将学到的经验归档为 skill，供以后参考。不要等用户提醒。.
oracle
Best practices for using the oracle CLI (prompt + file bundling, engines, sessions, and file attachment patterns).
agent-mode
Unified tool for managing agent LLM modes (add, remove, update, list, switch).
agento11y-prod-setup
Sets up production evaluation and guardrails for a DEPLOYED AI agent in Grafana Agent Observability, grounded in the agent's own code and its real ingested traffic. The judgment layer on top of the agento11y skill: it reads the agent's source (system prompt, tools, entrypoint) AND samples its live traffic via gcx…