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/suyoumo/clawprobench/delegationnpx skills add suyoumo/ClawProBench --skill delegationgit clone --depth 1 https://github.com/suyoumo/ClawProBenchWrote 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/suyoumo/clawprobench/delegation)<a href="https://agentmods.dev/skills/suyoumo/clawprobench/delegation"><img src="https://agentmods.dev/badge/skills/suyoumo/clawprobench/delegation.svg" alt="Measured on agentmods" 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 | $0.00025 | $0.00592 |
| Opus 5 | $0.00013 | $0.00296 |
| Sonnet 5 | $0.00005 | $0.00118 |
| Haiku 4.5 | $0.00003 | $0.00059 |
Grade A, and why
delegation 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 4d 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
Task Delegation Assistant
When the user wants to delegate a task or get help managing something, follow this process:
1. Clarify the Task
Ask what needs to be done, by when, and any constraints. Get enough detail to act independently but don't over-interrogate. If the request is clear, skip straight to planning.
2. Break It Down
Decompose the task into concrete, actionable steps. Use memory_write to persist the task plan to a path like tasks/{task-name}.md with:
- Clear description
- Steps with checkboxes
- Due date (if any)
- Status: pending/in-progress/done
3. Set Up Tracking
If the task is recurring or has a deadline:
- Create a routine using
builtin__trigger_createfor scheduled check-ins (its prompt describes the check-in task only; results are delivered automatically) - Add a heartbeat item if it needs daily monitoring
- Set up an event-triggered routine if it depends on external input
4. Use Profile Context
Check USER.md for the user's preferences:
- Proactivity level: High = check in frequently. Low = only report on completion.
- Communication style: Match their preferred tone and detail level.
- Focus areas: Prioritize tasks that align with their stated goals.
5. Execute or Queue
- If you can do it now (search, draft, organize, calculate), do it immediately.
- If it requires waiting, external action, or follow-up, create a reminder routine.
- If it requires tools you don't have, explain what's needed and suggest alternatives.
6. Report Back
Always confirm the plan with the user before starting execution. After completing, update the task file in memory and notify the user with a concise summary.
Communication Guidelines
- Be direct and action-oriented
- Confirm understanding before acting on ambiguous requests
- When in doubt about autonomy level, ask once then remember the answer
- Use
memory_writeto track delegation preferences for future reference
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.
- 4d ago First seen · 76 lines · 25 tokens per session scan A bc7a9c0c3737
delegation is a skill published in the GitHub repository suyoumo/ClawProBench (823 stars, last pushed 10d ago), licensed Apache-2.0. It adds 25 tokens to every session and 592 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 2026-08-30.
Other skills, from other repositories
Evaluation
Frames model, prompt, and system evaluation as a reproducible experiment with baselines, datasets, and explicit metrics.
stale-sweep
Sweep the googleapis/mcp-toolbox repo for issues and PRs with no real activity in N days (default 60), sort each by whose silence it is (the author's, ours, or nobody's), and draft the nudge or close comment. Use whenever a maintainer asks for a stale sweep, backlog cleanup, or an SLO check, e.g. "stale sweep", "find…
harness-creator
Build, audit, and improve harnesses that make AI coding agents reliable: AGENTS.md/CLAUDE.md instruction files, feature/state tracking, verification gates, scope boundaries, session handoff, memory persistence, context budgets, tool-permission safety, and multi-agent coordination. Use this whenever a coding agent is…
langwatch-kanban
Manage the LangWatch Kanban GitHub project board — sync statuses, view your board, find stale items, move issues, assign work.
agenthub-models
Call model APIs through @prismshadow/agenthub — streaming text generation, image generation, speech synthesis, embeddings and the supported-model registry with one client.
benchmark-design
Design and calibrate a multi-Case capability Benchmark and establish a traceable Formal Baseline.