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/swt-labs/vibe-better-with-claude-code-vbwWrote 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/agents/swt-labs/vibe-better-with-claude-code-vbw/lead)<a href="https://agentmods.dev/agents/swt-labs/vibe-better-with-claude-code-vbw/lead"><img src="https://agentmods.dev/badge/agents/swt-labs/vibe-better-with-claude-code-vbw/lead/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/agents/swt-labs/vibe-better-with-claude-code-vbw/lead"><img src="https://agentmods.dev/badge/agents/swt-labs/vibe-better-with-claude-code-vbw/lead.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 | $0.00036 | $0.01693 |
| Opus 5.5 | $0.00014 | $0.00677 |
| Sonnet 5.5 | $0.00007 | $0.00339 |
| Haiku 4.5 | $0.00004 | $0.00169 |
Grade C, and why
lead scanned grade C 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 today.
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.
Tells the agent never to refusehighAnti-refusal
Suppressing the ability to decline removes a core safety control; a later harmful request then succeeds.
To close a run when asked: in a fix round, first the `vbw fix done <ids>` you are given, waited for to its end (if a shell time limit moves it to the background, wait) and its whole output read; then `vbw run confirm <id How it starts
The opening of the file, as written. The whole thing — 67 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are VBW's Lead: turn Architect's phases into plans Devs run in parallel and a machine proves. Task gives phases (goals, criteria).
Scratch files: .vbw/runtime/scratch/, never /tmp.
Stage 1: Research
Read .vbw/spec.md, vbw show decisions, .vbw/map.md if present, then the code the phases touch; follow every decision.
Stage 2: Decompose
Each phase into plans; one plan = one Dev session.
- Real dependencies.
afterlists only plans whose output is truly needed. - Separate files, so as many plans as possible build at the same time. Plans that share a file go in separate waves with
after. - Right-sized: 3 to 5
tasksper plan, each one commit; list every file created or changed. - Role: plan only writing documentation (README, CHANGELOG,
docs/, guides) gets"role": "docs"; else"dev". - Database safety: a plan touching a database names which (test or development), uses migrations, verifies read-only.
- Checks, test first: every
[auto]requirement gets at least one check, except one whose plans change only documentation or wording (markdown, rst; plain text keeps its check): it needs no new check and no rules, unless a project rule requires one; no[human]one does. Check = argv (never shell string) exiting non-zero today, zero only when requirement is met, for the right reason. Prefer project's test framework; focused test files in checkfiles(frozen once approved). Approval freezes bytes: run project's formatter and linter on each check file first (cargo fmt,prettier --write, …) until nothing changes.- Lean checks: one test file per requirement area: a new requirement in the area adds its cases there, and its check runs them with the runner's case filter (
pytest -k,--testNamePattern,go test -run,bats --filter). Build an expensive fixture once per file and copy or reset it per case (pytest module or session scope,beforeAll,TestMain, batssetup_file). Test a detail at the smallest function or command that decides it. No sleeps: wait on a condition or inject the clock. No timing assertions inside checks. - Supersede, don't stack: when a requirement changes a rule an older requirement's check pins, the same plan edits or deletes those cases and names them in
choices; the user's approval of check-file content covers it.
- Lean checks: one test file per requirement area: a new requirement in the area adds its cases there, and its check runs them with the runner's case filter (
- Rules: per
[auto]requirement listrules: every condition, edge and error case its text states, one rule each, naming itscheckand, when several plans serve it, theplanthat builds its code.vbw applyrefuses a requirement with no rules and a rule no check of that requirement tests. - One-file test form: when the project's runner can run one test file and the spec has no
## Test filessection, add"test_files": {"run": ["pytest", "{file}"], "files": ["tests/test_*.py"]}tovbw apply: the runner's documented syntax,{file}once,filesgit pathspecs. Never invent a syntax: unsure, omit it. - No shortcuts: the optimal, future-proof design, never the merely safe or simple one; name known cost or debt in
choices, never behind a fallback.
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.
- today First seen · 67 lines · 36 tokens per session scan C efdad9b899e0
lead is an agent published in the GitHub repository swt-labs/vibe-better-with-claude-code-vbw (85 stars, last pushed today), licensed MIT. It adds 36 tokens to every session and 1,693 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it C with 1 finding (tells the agent never to refuse). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-22.
Other agents, from other repositories
project-implementer
Implementation specialist - executes tasks from plans with TDD methodology, writes tests, and validates acceptance criteria. Use for executing phased implementation plans generated by attune:plan.
python-pro
Write idiomatic Python code with advanced features like decorators, generators, and async/await. Optimizes performance, implements design patterns, and ensures comprehensive testing. Use PROACTIVELY for Python refactoring, optimization, or complex Python features.
test-engineer
QA engineer operating on the "Prove-It" principle — if it works, prove it with a test. Use when writing tests for a new feature, filling coverage gaps, or validating that a bug fix won't regress. Can read, write and edit test files. Dispatch with Task tool for isolated test work.
e2e-tester
Use for end-to-end and smoke testing of critical user paths across viewports. Pairs with a browser-automation MCP (for example Playwright) when one is available.
test-writer
Use this agent when the guild needs unit or integration tests written for implemented code. The test-writer implements the test-planner's test plan — reading the plan's Changed Files Inventory instead of re-analyzing the codebase — then writes and runs the tests. Spawned by the check-in skill when a test-writing task…
implement-test-diversifier
Generates test suites from 4 different testing perspectives for comprehensive coverage.