Borrowing it
Nothing to install: this file belongs to mtarcure/claude-vibe-squad. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/mtarcure/claude-vibe-squad/main/.claude/skills/lane-budget-estimation/SKILL.mdgit clone --depth 1 https://github.com/mtarcure/claude-vibe-squadWrote 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/mtarcure/claude-vibe-squad/lane-budget-estimation)<a href="https://agentmods.dev/skills/mtarcure/claude-vibe-squad/lane-budget-estimation"><img src="https://agentmods.dev/badge/skills/mtarcure/claude-vibe-squad/lane-budget-estimation/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/mtarcure/claude-vibe-squad/lane-budget-estimation"><img src="https://agentmods.dev/badge/skills/mtarcure/claude-vibe-squad/lane-budget-estimation.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00077 | $0.00596 |
| Opus 5 | $0.00039 | $0.00298 |
| Sonnet 5 | $0.00015 | $0.00119 |
| Haiku 4.5 | $0.00008 | $0.00060 |
Grade A, and why
lane-budget-estimation 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 8d 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 — 40 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Lane Budget Estimation
House procedure for sizing a dispatch to the two hard budgets a lane runs under: the wall clock and the assembled-prompt ceiling. Over-scoping does not degrade gracefully — it dies at the wall with nothing to show. Verify the numbers against the live preflight, not from memory.
The prompt ceiling is on the ASSEMBLED prompt
- The limit is
TRUSTED_LAUNCH_PROMPT_LIMIT = 40960bytes (dispatch_context_builder.py), applied to the packet PLUS injected briefing and absolute paths, not the packet body alone. Bothdispatch_preflight.evaluate_packet()andbuild_context()enforce the same constant. - Do not estimate from a remembered packet size — injected context and absolute path length vary by
lane and checkout. Run
bin/send-task.sh <packet-file> --dry-runand readassembled_prompt_bytesagainstassembled_prompt_limitfrom the preflight; shorten the body if it refuses.
Scope to one wall
- The wall comes from
timeout_budget_for_mode(): a flat 2700s forproject, 3600s forbounty. It is a backstop, not a per-packet dial you raise. - A deliverable that cannot finish in one wall is two packets, not one. Sequence them; the second reads the first's committed output (an uncommitted edit is invisible to the next dispatch).
- A lane that must read a large corpus AND produce a large artifact is two packets — the read burns the wall the artifact needs.
Name what the lane must NOT do
- Unbounded exploration is how a wall gets hit with nothing landed. Put explicit out-of-scope bounds in the packet: which trees not to read, which analyses to skip, when to stop and write the partial.
- Tell the lane to land what is complete and write the artifact before the wall rather than chase completeness past it — a truthful partial beats a kill with no envelope.
Size the read, then the write
- Estimate the read set first (
scope-estimationsizes the corpus), because a read that cannot fit the wall guarantees the write never starts. - Round-trips, not tool calls, are the cost unit that eats a wall: prefer whole-file reads and tool-level exclusion globs over paging, and say so in the packet if the lane tends to page.
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.
- 8d ago First seen · 40 lines · 77 tokens per session scan A 7ad1f484e5a1
lane-budget-estimation is a skill published in the GitHub repository mtarcure/claude-vibe-squad (148 stars, last pushed 2d ago), licensed MIT. It adds 77 tokens to every session and 596 once invoked, about $0.0004 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 skills, from other repositories
gh-find-prs
Survey open Codewhale PRs and triage each for mergeability and disposition against the real landing branch.
contributor-onboarding
Help a new contributor get productive on this checkout - inspect sync state against main, build, run the repository's exact verification gate, and produce a local what's-new digest. Never fetches, pulls, or modifies a dirty tree on its own. Explicit-only.
codew-release-qa-sweep
Use before claiming Codewhale release work is done: run the full gate sweep and list the manual QA targets.
gh-file-issue
Use when filing a new Codewhale GitHub issue: turn a bug or idea into a well-formed, actionable issue with repro, acceptance criteria, labels, and milestone.
gh-treasure-hunt
Hunt the issue/PR queue for highest value-over-risk wins: clean focused community PRs, already-implemented issues to close, safe quick-fixes.
skill-installer
Install, update, trust, or inspect Codewhale skills from GitHub or local skill folders. Use when the user asks for available skills or wants a community skill installed.