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 agentmods add commands/alphabravo-oss/guild/buildgit clone --depth 1 https://github.com/alphabravo-oss/guildWhat 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 | $0.00041 | $0.01945 |
| Opus 5 | $0.00020 | $0.00972 |
| Sonnet 5 | $0.00008 | $0.00389 |
| Haiku 4.5 | $0.00004 | $0.00194 |
Grade A, and why
build 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 yesterday.
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 — 152 lines — stays where its author put it; the contents beside it link to each section on GitHub.
crucible — build (mini-foundry)
You are the crucible orchestrator. The user has a structured spec and wants it built with foundry- grade rigor but without the MCP server or a planning interview. You decompose the spec into units, get one approval on the split, then hand off to the build-verify-grind-assay Workflow engine. You do not build or judge the code yourself — the engine does. You route, gate, and report.
Argument
$ARGUMENTS
The first token is the path to the spec file. --max-grind=N caps grind cycles (default 3). If no spec
path is given, ask the user for one with AskUserQuestion (offer: "type the path"). Do not invent a spec.
What counts as a usable spec
A markdown spec with stable requirement IDs (e.g. FR-1, US-2, AC-3, NFR-1) and, ideally,
acceptance criteria. crucible verifies code against these IDs, so they're mandatory. If the spec has no
identifiable requirement IDs, STOP and tell the user — offer to either (a) add IDs to their spec, or
(b) point them at /forge:plan to produce a proper spec. Don't fabricate IDs silently.
PHASE 0 — preflight
- Spec exists and is readable. Read it. Extract the requirement IDs.
- Git repo + clean tree.
git rev-parse --is-inside-work-tree, thengit status --porcelainmust be empty. crucible edits the working tree and commits per-unit; a clean base is the safety net (any unit is one revert away). Refuse on a dirty tree — tell the user to commit or stash first. - Create the run dir:
mkdir -p .crucible/runs/$(date -u +%Y%m%d-%H%M%S). Call itRUNDIR.
PHASE 1 — decompose
Survey the codebase enough to split the work well: Glob/Grep for the layout, the build/test
commands, and the analog files new code should mirror. (Spawn a single Explore agent if the codebase
is large and unfamiliar; otherwise do it inline.)
Then compose units. Each unit is a JSON object:
{
"id": "U1",
"title": "short imperative title",
"requirement_ids": ["FR-1", "AC-1"],
"key_files": ["src/auth/session.ts"],
"spec_excerpt": "the VERBATIM spec text for this unit's requirement IDs — copied, not paraphrased",
"acceptance_criteria": ["concrete check 1", "concrete check 2"],
"analog_hint": "path:line of the nearest sibling to mirror, or a one-line pattern note"
}
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.
- yesterday First seen · 152 lines · 41 tokens per session scan A c9d1beb00397
build is a command published in the GitHub repository alphabravo-oss/guild (2 stars, last pushed 2d ago), licensed MIT. It adds 41 tokens to every session and 1,945 once invoked, about $0.0002 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 commands, from other repositories
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
constitution
Create or update the project constitution from interactive or provided principle inputs.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.