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 atman-33/workhub --skill task-handoffgit clone --depth 1 https://github.com/atman-33/workhubWrote 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/atman-33/workhub/task-handoff)<a href="https://agentmods.dev/skills/atman-33/workhub/task-handoff"><img src="https://agentmods.dev/badge/skills/atman-33/workhub/task-handoff/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/atman-33/workhub/task-handoff"><img src="https://agentmods.dev/badge/skills/atman-33/workhub/task-handoff.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 | $0.00062 | $0.01175 |
| Opus 5.5 | $0.00025 | $0.00470 |
| Sonnet 5.5 | $0.00012 | $0.00235 |
| Haiku 4.5 | $0.00006 | $0.00118 |
Grade A, and why
task-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 27d 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 — 92 lines — stays where its author put it; the contents beside it link to each section on GitHub.
task-handoff — Continue here, or hand the task to a fresh session
A task filed mid-session invites rolling straight into it. The owner's default
runs the other way: a follow-up gets its own session, with a clean context and
its own row in the session list. This skill turns that into an explicit call —
a recommendation with its reason, every time — and carries out the handoff.
For an ordered list of tasks to run back to back, use task-chain instead;
to decide the order and parallelism of many tasks first, use task-orchestrate.
Steps
-
Recommend continue or hand off. Read the task file's frontmatter and
## Description. This skill is for a task an agent is meant to work (todo, an agent assignee); for aninboxidea or a task assigned tome, the filing is the whole job and this skill is done — unless the owner asked for the handoff themselves. Recommend a fresh session when any of these hold:- it is its own unit of work — a different milestone or backlog item from the task this session is on;
- it has
confirm: true, so its plan needs the owner's approval; - it works in a different area or repository from this session;
- this session's context is already large;
- this session's own task is still open —
task-reporthas not run.
Recommend continuing here only when all of these hold: the change is small (a few places), it leans on context this session holds and no note records, and the owner has said to carry on. When the call is close, recommend a fresh session.
Put the recommendation and its one-line reason to the owner, and wait for their answer. On "continue here", run
task-startin this session; this skill is done. -
Pick the chip's start mode. Applies when this session has the desktop app's
spawn_tasktool (Claude Desktop). Starting a chip asks the owner whether to run it locally or in a git worktree. Two different worktrees are in play; keep them apart:- The chip's worktree is made of this session's working directory, on a
claude/...branch. In the vault, that is a worktree of the vault. worktree: truein a task's frontmatter makestask-startbuild a worktree of the target repository ontask/<task-id>. It is independent of the chip's start mode, and both can apply to one task.
- The chip's worktree is made of this session's working directory, on a
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.
- 27d ago Changed · +9 lines ec0182234c9b
- 28d ago Changed · +2 lines 40175f123e43
- 29d ago Changed · +1 lines 51f71f0384f4
- 1mo ago First seen · 80 lines · 62 tokens per session scan A 82264f36e5a2
task-handoff is a skill published in the GitHub repository atman-33/workhub (3 stars, last pushed yesterday), licensed MIT. It adds 62 tokens to every session and 1,175 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-09-30.
Other skills, from other repositories
expo-launchpad-status
Show the current expo-launchpad pipeline state — phase, round, next role, pause reason — read-only.
tasknotes
This skill should be used when the user asks to 'create a task', 'add a TODO', 'plan a sprint', 'organize tasks', 'track time', 'setup Kanban', 'create task from code', 'sync tasks with sprint', 'pomodoro session', 'start pomodoro', 'time estimate', or mentions TaskNotes, task management, or project planning in…
project-consolidation
Safely consolidate drifted/duplicate project folders in the memex vault — cwd-fragment folders (e.g. Apps-arena, GitHub-loom), duplicate/split projects (one project's content scattered across two folders), and name-mismatch folders. Trigger on "consolidate project folders", "merge duplicate projects", "fix project…
issue-hygiene
How to write an issue somebody can act on, how to triage a pile of them, and the label vocabulary that stays useful. Use when filing, triaging, labelling, closing or linking issues.
aetox-slice
Decomposes a technical specification or plan into vertical tracer-bullet tickets. Calculates directed dependency graphs (DAG), schedules prefactoring first, applies expand-contract sequencing for wide refactors, quizzes the user on ticket granularity, and emits atomic per-ticket files sized for optimal agent context…
roadmap-update
Produit un document de roadmap dans output/ en croisant l'état actuel des epics beads avec les décisions et priorités documentées dans le wiki. Applique les règles anti-lissage — ne jamais compléter avec des connaissances d'entraînement, citer le wiki pour chaque affirmation de scope ou de priorité. Triggers — mettre…