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 yunsii/wezdeck --skill cross-repo-delegategit clone --depth 1 https://github.com/yunsii/wezdeckWrote 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/yunsii/wezdeck/cross-repo-delegate)<a href="https://agentmods.dev/skills/yunsii/wezdeck/cross-repo-delegate"><img src="https://agentmods.dev/badge/skills/yunsii/wezdeck/cross-repo-delegate/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/yunsii/wezdeck/cross-repo-delegate"><img src="https://agentmods.dev/badge/skills/yunsii/wezdeck/cross-repo-delegate.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 | $0.00111 | $0.02786 |
| Opus 5.5 | $0.00044 | $0.01114 |
| Sonnet 5.5 | $0.00022 | $0.00557 |
| Haiku 4.5 | $0.00011 | $0.00279 |
Grade A, and why
cross-repo-delegate 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 20d 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 — 214 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Cross-repo delegate (work tickets)
Who runs: the coding agent (Host TUI / OpenClaw), not the human.
Never tell the user to copy-paste run.sh as the primary path.
Single source: this directory (SKILL.md + run.sh + test.sh). Other paths
are symlinks from scripts/dev/link-platform-skills.sh.
Skill id: cross-repo-delegate. Short CLI on PATH (optional): delegate.
Three modes (hard rule)
| Mode | User intent | You do | Must NOT |
|---|---|---|---|
| 1. 主会话创建工单 | 提单 / 跨仓委托 / 只交票 | create（无 --run） |
自动 run / worktree / worker |
| 2. 主会话认领并开发 | 有没有我的单；认领；认领后改 | inbox → claim → show → 在当前 TUI/cwd 开发 |
run / create --run / 建 delegate-* worktree |
| 3. 主会话建单并委托开发 | 明确说后台跑 / 派工人 / 委托开发 | create … --run 或已有单上 run --phase auto |
把 Mode 2 的认领误当成 Mode 3 |
claim 默认 session（owner=human）。Headless worker 才用 claim --as-worker（由 run 内部调用）。
认领 ≠ 派工人。 用户在 TUI 认领后，由当前主会话读票、调研、改代码、challenge / close。
Concrete repo names live only in ~/.agent/tickets/config.yml (and the user’s words), not in this skill’s trigger text.
Routing: ticket vs multi-dir mount
Default for 跨仓委托 / 认领 / 挑战 / 关单: this skill (Modes 1–3).
Product multi-dir mounts (Claude --add-dir, Codex multi-folder, Cursor multi-root, parent-folder workspaces) are a tactical way to keep one semantic edit coherent across checkouts. They do not replace the ticket protocol (ownership, observed/assumed, challenge/reply, allowlist, audit).
| User signal | You do |
|---|---|
| 提单 / 跨仓委托 / 有没有我的单 / 认领 / 挑战假设 / 关单 / 派工人 | Modes 1–3 below |
| Same intent must land in 2+ repos in one sitting (shared type rename, co-edited contract) | Allow a short product multi-dir session; do not invent a second ticket store; prefer tickets when ownership or challenge is needed |
| Source-repo agent wants to “keep context” by writing the allowlisted target as standing practice | Refuse that framing — file/claim on the target (Mode 1/2) or Mode 3 worker on a target worktree |
What ships with it
13 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.
- config.example.yml 404 B
- delegate 338 B
- lib/common.sh 8.0 KB runs code
- lib/lifecycle.sh 14 KB runs code
- lib/ticket_fs.py 13 KB runs code
- lib/worker.sh 15 KB runs code
- lib/worktree.sh 2.6 KB runs code
- prompts/implement.md 1019 B
- prompts/research.md 1.7 KB
- README.md 2.5 KB
- run.sh 31 KB runs code
- templates/tickets-README.md 1.8 KB
- test.sh 15 KB runs code
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.
- 20d ago Changed f60008de82cf
- 22d ago Changed · +18 lines 26ff19931377
- 29d ago Changed · +28 lines 107a03846333
- 1mo ago Changed · +4 lines 7755b643b8ea
- 1mo ago First seen · 164 lines · 111 tokens per session scan A 4cba22f173b5
cross-repo-delegate is a skill published in the GitHub repository yunsii/wezdeck (5 stars, last pushed yesterday), licensed MIT. It adds 111 tokens to every session and 2,786 once invoked, about $0.0004 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 2026-09-13.
Other skills, from other repositories
paperclip
Use for Paperclip-managed tasks and heartbeats: reading task context, delivering task documents or files, updating completion or blockers, coordinating or delegating work, and following company governance. Includes control plane API operations for assignments, comments, approvals, and routines.
paperclip-task-bridge
Create, comment on, update, and list Paperclip tasks from Hermes using scoped Paperclip API credentials.
issue-triage
Triage Paperclip inbox issues that are stale, blocked, in-review, or assigned-but-not-progressing, and decide a single next action per issue (resume, reassign, unblock, escalate, or close).
gh-file-issue
Use when filing a new Codewhale GitHub issue: turn a bug or idea into a well-formed, actionable issue with repro, acceptance criteria, labels, and milestone.
gh-treasure-hunt
Hunt the issue/PR queue for highest value-over-risk wins: clean focused community PRs, already-implemented issues to close, safe quick-fixes.
review-my-team
Review a connected Paperclip team's agents, tasks, blockers and pending approvals when the user asks for a team summary or what needs attention.