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.
git clone --depth 1 https://github.com/travisjneuman/.claudeWrote 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/commands/travisjneuman/.claude/start-task)<a href="https://agentmods.dev/commands/travisjneuman/.claude/start-task"><img src="https://agentmods.dev/badge/commands/travisjneuman/.claude/start-task/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/commands/travisjneuman/.claude/start-task"><img src="https://agentmods.dev/badge/commands/travisjneuman/.claude/start-task.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 | $0.00022 | $0.00729 |
| Opus 5 | $0.00011 | $0.00365 |
| Sonnet 5 | $0.00004 | $0.00146 |
| Haiku 4.5 | $0.00002 | $0.00073 |
Grade B, and why
start-task scanned grade B with 1 finding 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 9d 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
- For specialized domains: `find ~/.claude/plugins/marketplaces -name "SKILL.md" | xargs grep -li "<keyword>"` How it starts
The opening of the file, as written. The whole thing — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Start Task - Universal Intelligent Router v3.2
Routes ANY prompt to optimal execution by loading resources on-demand. Skills and agents auto-match from descriptions. CLAUDE.md routes docs/checklists.
Execution Protocol
Step 0: ULTRATHINK (Default)
Apply maximum reasoning depth for non-trivial tasks. Skip only for: typos, single-line fixes.
Step 1: Environment Context & Research
Current date from environment: Use Today's date value for temporal reasoning.
Research triggers (auto-detect):
- "latest", "current", "recent", "new", "today"
- Any year >= current year from environment
- Version queries ("React 19", "Node 22")
- "best way to", "recommended", "is there a library"
- "investigate", "research", "find information about", "compare approaches"
If research needed: Use WebSearch immediately. For complex research: Load docs/reference/workflows/research-methodology.md.
Parallel tool usage: When multiple independent searches/reads needed, make ALL calls in single message.
Step 2: Memory Check
Query persistent memory for context relevant to: {{task_description}}
Step 3: Active GSD Project Check
[ -f .planning/STATE.md ] && echo "ACTIVE_GSD_PROJECT"
If active: Invoke /gsd:progress and EXIT.
Step 4: Domain Detection
Scan {{task_description}} and load the matching domain router file:
- Meta/Admin → Read
commands/router/domains-meta.md→ May EXIT immediately - Decision → Read
commands/router/decision-frameworks.md→ EXIT after framework - Scientific → Read
commands/router/domains-scientific.md - Development → Read
commands/router/domains-development.md - Business → Read
commands/router/domains-business.md - Creative → Read
commands/router/domains-creative.md
Extract from the domain file: skill to invoke, agent to spawn, contextual rules to read.
Step 5: Route Selection
Read commands/router/routing-logic.md for complexity scoring.
- Score >= 5 →
/gsd:new-project - Score 3-4 →
EnterPlanMode(or suggest Auto-Claude for well-defined features) - Score 1-2 →
EnterPlanMode - Score <= 0 → Execute directly
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.
- 9d ago First seen · 83 lines · 22 tokens per session scan B 318290ed2da0
start-task is a command published in the GitHub repository travisjneuman/.claude (97 stars, last pushed 7d ago), licensed MIT. It adds 22 tokens to every session and 729 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other commands, from other repositories
api-gen
Generate REST API endpoints with routes, validation, error handling, and tests.
component-gen
Generate React or Vue components with prop types, styling, accessibility, and tests.
performance-audit
Find performance bottlenecks across the stack — queries, rendering, bundle size, memory.
debug
Systematic debugging — reproduce, isolate, fix, verify, and add regression test.
doc-gen
Generate documentation — JSDoc, docstrings, README sections, or API docs.
migrate
Migrate between framework versions or languages with an audited step-by-step plan.