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/triumsebas/unlimited-mcp/delegatenpx skills add triumsebas/unlimited-mcp --skill delegategit clone --depth 1 https://github.com/triumsebas/unlimited-mcpWrote 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/triumsebas/unlimited-mcp/delegate)<a href="https://agentmods.dev/skills/triumsebas/unlimited-mcp/delegate"><img src="https://agentmods.dev/badge/skills/triumsebas/unlimited-mcp/delegate.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 5.1 | $0.00000 | $0.01402 |
| Opus 5 | $0.00000 | $0.00701 |
| Sonnet 5 | $0.00000 | $0.00280 |
| Haiku 4.5 | $0.00000 | $0.00140 |
Grade A, and why
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 6d 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 — 144 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Skill: delegate
Opinionated, Claude-Code-specific patterns on top of the unlimited-mcp MCP
tools. Requires the unlimited-mcp server to be connected.
Read the shared reference first
This skill does not restate the tool reference. The canonical guidance —
queue selection, sync vs background, JobResult, safety, workspace presets,
run_command vs run_shell, run_and_summarize vs smolagents, the timeout
multiplier table, clarify_rounds / await_worker_questions, sub-task
grouping, local-GPU agent discipline, the before-task checklist, and the
new-agent verification tests — lives in AGENTS.md at the repo root.
Read AGENTS.md for any of those. This file only adds the named delegation
wrappers and the Claude-Code-only behaviours below.
The three delegation patterns
delegate.now — sync, small tasks (< 30s)
Use when the task is short and you need the diff inline this turn. Never for agent coding tasks (those always go background).
1. add_allowed_root(cwd)
2. r = delegate_to_agent(agent_name, prompt=prompt, cwd=cwd,
workspace='safe_dev', queue='local')
3. r = await_job(r.job_id)
4. Return r.summary and r.branch (if any)
delegate.fire_and_forget — background, long tasks
Any agent coding task, anything > 1 min, or multiple parallel tasks. Uses
queue="ts" so the job survives MCP restarts and shows in the inbox.
1. add_allowed_root(cwd)
2. job = submit_task(agent_name=agent_name, prompt=prompt, cwd=cwd,
queue='ts', tag=session_tag)
3. Tell the user: "Job submitted: {job.job_id}."
4. Return immediately — recover later via list_jobs(tag=session_tag).
Pass a consistent tag (date or task name) so a new session can find the job.
delegate.await — background, then report on request
1. add_allowed_root(cwd)
2. job = submit_task(..., queue='ts', idempotency_key=key)
3. On user request: r = get_job_result(job.job_id)
- running → tell the user it's still running
- completed → show r.summary and r.diff_ref
- failed → show r.error and r.summary
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.
- 6d ago First seen · 144 lines · 0 tokens per session scan A 430b310192a3
delegate is a skill published in the GitHub repository triumsebas/unlimited-mcp (2 stars, last pushed 3mo ago), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 1,402 tokens. 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 2026-08-31.
Other skills, from other repositories
snip
You are an expert at writing declarative YAML filters for snip, a CLI proxy that reduces LLM token consumption by filtering shell output.
platform_management
/workspace-root = volume mount of parent dir.
github_read
Read access: repos, issues, PRs, files, branches, workflows, runs.
gitlab_read
Read access: projects, issues, MRs, files, pipelines, members.
schedule_manage
Tool: schedulemanage. All ops use action field.
agent_spawn
Spawn specialized sub-agents → handle independent workstreams in parallel.