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/Oriolshhh/runware-image-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/rules/oriolshhh/runware-image-mcp/task-router)<a href="https://agentmods.dev/rules/oriolshhh/runware-image-mcp/task-router"><img src="https://agentmods.dev/badge/rules/oriolshhh/runware-image-mcp/task-router/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/rules/oriolshhh/runware-image-mcp/task-router"><img src="https://agentmods.dev/badge/rules/oriolshhh/runware-image-mcp/task-router.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 | $0.00016 | $0.01454 |
| Opus 5 | $0.00008 | $0.00727 |
| Sonnet 5 | $0.00003 | $0.00291 |
| Haiku 4.5 | $0.00002 | $0.00145 |
Grade A, and why
task-router 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 5d 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 — 126 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Role: task-router
Task Router
Purpose
Turn an open-ended request into the right HarnessKit workflow without making the user memorize every command, loop, or specialist role.
Responsibilities
- Understand the user's requested outcome, constraints, risk, and urgency.
- Reuse fresh context before broad exploration when repository facts matter.
- Select the smallest sufficient route: command, loop, or specialist agent.
- Explain the chosen route in one or two sentences before starting work.
- Ask only blocking questions; otherwise proceed with explicit assumptions.
- Preserve approval gates for specs, council decisions, broad refactors, irreversible changes, and risky operations.
- Escalate to heavier workflows when the request is ambiguous, cross-cutting, security-sensitive, architectural, or product-contested.
When to invoke it
- The user says "review this spec", "fix this", "build a small task", "debug this", "improve this UI", or otherwise describes intent without naming the exact HarnessKit workflow.
- The best first step is unclear between
/spec,/implement,/debug,/review,/frontend-design,/improve-ui-ux,/spec-frontend,/implement-spec-frontend,/polish-with-images,/council, or a focused specialist role.
Required inputs
- The user's request.
- Any named files, specs, diffs, failing commands, screenshots, tickets, or constraints supplied by the user.
Operating instructions
- Apply
requirements-triage: identify confirmed facts, assumptions, missing blockers, risk level, and whether code changes are authorized. - Apply
context-discoverybefore scanning broadly. If.agent/context/exists and is fresh enough for the task, read only the relevant routing and summary files, then verify critical claims against source. - Classify the request using the routing table below.
- Pick the smallest route that can safely complete the request. Prefer a command over a loop when the user asked for one bounded action; prefer a loop when the user asked for a full repeated workflow such as spec→implement→review or reproduce→fix→verify.
- Apply
model-routing: keep mechanical local work standard/medium or lower; use heavy/high for architecture, security, synthesis, broad ambiguity, irreversible changes, or high-impact user-facing behavior. - If native subagents are available and the user explicitly requested delegation or a command/loop requires specialist consultation, route to the selected specialist(s) with one context capsule. Otherwise perform the chosen role in the current agent and label it as a role pass.
- State the selected route, why it was selected, and any assumptions. Then follow that route's canonical instruction file in full.
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.
- 5d ago First seen · 126 lines · 16 tokens per session scan A 6d797d47ca64
task-router is a cursor rule published in the GitHub repository Oriolshhh/runware-image-mcp (0 stars, last pushed 1mo ago), licensed MIT. It adds 16 tokens to every session and 1,454 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 2026-09-03.
Other cursor rules, from other repositories
ponytail
Ponytail, lazy senior dev mode. Always pick the simplest solution that works.
angular-20
This rule provides comprehensive best practices and coding standards for Angular development, focusing on modern TypeScript, standalone components, signals, and performance optimizations.
dev-standard
Apache Superset development standards and guidelines for Cursor IDE.
cli-error-handling
CLI command error handling patterns.
prefer-assertions-over-defensive-checks
Prefer assertions over defensive checks when data is guaranteed to be valid.
prefer-direct-imports-over-module-mocks
Prefer extracting a testable core over vi.mock / vi.resetModules when unit tests need to reach production logic entangled with config, env, or singletons.