build

A development-phase command that prepares the relevant project context and sends work to one or more specialist work lanes.

In plain words
What is it for?
Use it to run the development phase or rerun one specific lane, such as a backend lane.
Why use it?
It organizes implementation work so each specialist receives the context needed for its part of the task.

Command

Install

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.

agentmods
npx agentmods add commands/lookatitude/guild/build
Clone the repo
git clone --depth 1 https://github.com/lookatitude/guild
Per session 51 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,746 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00051 $0.02746
Opus 5 $0.00026 $0.01373
Sonnet 5 $0.00010 $0.00549
Haiku 4.5 $0.00005 $0.00275

Measured 2d ago against content hash ce034aa6d68c, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

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 2d 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.

commands/build.md · 217 lines

How it starts

The opening of the file, as written. The whole thing — 217 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/guild:build — phase: Development

The Development phase entrypoint. Assembles per-specialist context and dispatches lanes. The backend (agent_mode) is resolved once at command intake via the run-start preflight snapshot — it is not selected per-phase. snapshot.effective.agent_mode (team / agent / subagent) drives lane dispatch; execute-plan reads it from the locked-in snapshot rather than re-resolving. An optional positional [lane-id] re-runs a single lane.

Verb↔phase edge: /guild:build → Development. Phase concept binding: lifecycle/phase-entrypoints.md · lifecycle/lifecycle-overview.md.

Contract binding

Before producer work begins, this phase resolves the frozen guild.phase_entry.v1 contract. The run executes under the task_run.autonomy_policy recorded in the approved plan (the autonomy contract was set at plan approval; no new gate).

Usage

/guild:build
/guild:build lane-backend-001

All five global flags + --dry-run apply.

Args & local flags

  • Args: [lane-id] (re-run one lane)
  • Local flags:
    • --model-tier=cheap|mid|powerful — top-precedence tiering override; forces every lane in this run to the given model tier regardless of per-lane plan overrides or the models: settings block. Precedence: --model-tier > per-lane model_tier: > settings.json models: > built-in. Ref: docs/v2/config-surfaces.html §5. DRIFT-ANALYSIS CMD-008.

Gates (default)

  • Autonomy contract (set at plan approval) A
  • G-lane review A (per lane, via guild:review-broker; policy-gated — fires when risk ≥ high OR --review=cross OR project config requires it)
  • Destructive / network ops I always (immutable always-ask hard set — never relaxed by --auto-approve)

Output artifact

.guild/context/<run-id>/<specialist>-<task-id>.md (per-lane context bundles), .guild/runs/<run-id>/handoffs/*.md (guild.handoff_receipt.v1), .guild/runs/<run-id>/assumptions.md, .guild/runs/<run-id>/review.md, .guild/runs/<run-id>/review/G-lane:<task-id>/ (G-lane broker trail per lane), .guild/runs/<run-id>/agent-bus/events.ndjson, .guild/runs/<run-id>/verify.md (guild:verify-done), changed files.

Read the full file on GitHub · 217 lines

Changes

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.

  1. 2d ago First seen · 217 lines · 51 tokens per session scan A ce034aa6d68c

Subscribe to this mod's changes

build is a command published in the GitHub repository lookatitude/guild (7 stars, last pushed 2d ago), licensed MIT. It adds 51 tokens to every session and 2,746 once invoked, about $0.0003 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.