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 wan-huiyan/agent-traffic-control --skill credit-stall-mid-orchestration-revive-collisiongit clone --depth 1 https://github.com/wan-huiyan/agent-traffic-controlWrote 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/wan-huiyan/agent-traffic-control/credit-stall-mid-orchestration-revive-collision)<a href="https://agentmods.dev/skills/wan-huiyan/agent-traffic-control/credit-stall-mid-orchestration-revive-collision"><img src="https://agentmods.dev/badge/skills/wan-huiyan/agent-traffic-control/credit-stall-mid-orchestration-revive-collision/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/wan-huiyan/agent-traffic-control/credit-stall-mid-orchestration-revive-collision"><img src="https://agentmods.dev/badge/skills/wan-huiyan/agent-traffic-control/credit-stall-mid-orchestration-revive-collision.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 | $0.00303 | $0.03297 |
| Opus 5 | $0.00151 | $0.01648 |
| Sonnet 5 | $0.00061 | $0.00659 |
| Haiku 4.5 | $0.00030 | $0.00330 |
Grade A, and why
credit-stall-mid-orchestration-revive-collision 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 12d 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 — 265 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Credit-stall mid-orchestration: revive-collision recovery
Problem
You're orchestrating N parallel subagents (Task tool with run_in_background: true),
each working on independent tracks (e.g., one PR per track). Mid-run, an Anthropic
billing/credit failure or other transient harness outage hits — all in-flight subagents
freeze simultaneously. The user fixes billing, reports back, and asks you to continue.
The trap: stalled subagents are NOT dead — they're suspended at the harness layer. When API access is restored, the harness can auto-resume them from the exact tool-call boundary they froze on. If you naively relaunch "resume" agents to pick up where you think the originals stopped, you end up with TWO agents racing on:
- The same git branch (one commits + pushes, the other rebases on top of stale state)
- The same PR (duplicate review comments, conflicting label changes)
- The same files (one Edit lands, then the other Edit overwrites)
Symptoms when this happens: PR has two "no issues found" review comments, force-push storms as both agents rebase, agent A reports "PR already merged" while agent B is still polling CI for that same PR.
Severity ladder (observed in S156, 2026-05-08, all three originals revived ~2hr post-stall):
- Best case (Track A original): revives mid-implementation, notices the merged PR via subsequent harness signals, sends a status update, terminates. Resume agent had pivoted to value-add (PR #571) under a state-aware brief. No debris.
- Middle case (Track C original): revives, runs
git diff origin/main HEAD -- <files>, finds zero diff on feature files, recommends "do not file PR / delete branch", terminates. No PR debris, just a status-report task-notification. - Worst case (Track D original): revives, FILES a full duplicate PR against the already-merged branch (recreating the deleted remote branch via push), THEN runs its diff check, detects the collision, posts a "close as superseded" comment to its own duplicate PR, terminates. Debris: 1 open PR + recreated remote branch → orchestrator must close + delete.
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.
- 12d ago First seen · 265 lines · 303 tokens per session scan A 99501005fecf
credit-stall-mid-orchestration-revive-collision is a skill published in the GitHub repository wan-huiyan/agent-traffic-control (3 stars, last pushed yesterday), licensed MIT. It adds 303 tokens to every session and 3,297 once invoked, about $0.0015 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 2026-08-31.
Other skills, from other repositories
codex-coplan
Bring in an external model (Codex) to co-develop a plan — both models draft independently, align on disagreements, and merge into one plan. Load this only when the user explicitly names Codex (e.g. "let's discuss this with Codex," "what does Codex think about this approach") — don't auto-trigger just because a plan…
codex-orchestrator
Delegate execution tasks to Codex (Claude plans and reviews). Load this only when the user names Codex explicitly — no matter how big the task is, if the user hasn't mentioned Codex, do it yourself.
good-bad-ugly
Use when the user asks to "orchestrate", "conduct", invoke "good bad ugly" / "gbu", or tackle a large multi-part coding task (audits, refactor sweeps, multi-feature sessions, "clean this up and fix everything") — anything too big for one linear pass. Runs the tiered orchestration workflow: The Good (brain) plans and…
orchestrate
Use only when the user explicitly types /orchestrate:orchestrate to decompose a large task, spawn a tree of parallel worker/subplanner/verifier subagents, and collect structured handoffs. Do not invoke autonomously.
api-testing
HTTP API testing for TypeScript (Supertest) and Python (httpx, pytest). Test REST APIs, GraphQL, request/response validation, authentication, and error handling.
bun-test-coverage
Use for test coverage with Bun, --coverage flag, lcov reports, thresholds, and CI integration.