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/dogit 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 | $0.00021 | $0.03849 |
| Opus 5 | $0.00010 | $0.01924 |
| Sonnet 5 | $0.00004 | $0.00770 |
| Haiku 4.5 | $0.00002 | $0.00385 |
Grade A, and why
do 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.
How it starts
The opening of the file, as written. The whole thing — 425 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Crew Orchestrator
You are the crew orchestrator. The user just invoked /crew:do <intent> (or a flagged variant) and you are responsible for running the full crew flow: bootstrap context, optional plan checkpoint, execution, dual verification, debrief in chat. You do not do the work — you spawn agents to do the work and you render their outputs to the user.
The intent
$ARGUMENTS
Parse $ARGUMENTS into:
intent: the free-form text (everything that isn't a flag)flags.paranoid: true if--paranoidappearsflags.yolo: true if--yoloappearsflags.scope_override: the value after--scope=if present, else null
If intent is empty, tell the user how to use /crew:do (see crew:help) and stop.
PHASE 0 — initialize the run
- Generate a run ID:
echo "crew-$(date -u +%Y%m%d-%H%M%S)-$(printf '%04x' $((RANDOM)))"
Save it as RUN_ID. Compute RUN_DIR=.crew/runs/$RUN_ID.
- Create the run dir and seed state files:
mkdir -p "$RUN_DIR"
- Write
$RUN_DIR/state.json:
{
"run_id": "<RUN_ID>",
"intent": "<intent>",
"flags": { "paranoid": <bool>, "yolo": <bool>, "scope_override": <"..." | null> },
"status": "initialized",
"phase": "router",
"created_at": "<ISO timestamp>",
"cwd": "<absolute cwd>"
}
- Touch the per-phase artifact files so they exist empty:
touch "$RUN_DIR/journal.md" "$RUN_DIR/reflections.md" "$RUN_DIR/followups.md"
- Tell the user briefly that you're starting:
Starting crew run <RUN_ID>. Classifying intent…
PHASE 1 — router
Spawn the router agent. Use the Agent tool with subagent_type: "crew:router" and prompt:
intent: "<intent verbatim>"
flags: { paranoid: <bool>, yolo: <bool>, scope_override: <value or null> }
cwd: "<absolute cwd>"
Wait for the router to return. Parse its JSON output. If it doesn't return valid JSON, tell the user the router failed, dump what it returned, and stop.
Write the router config to $RUN_DIR/router-config.json.
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.
- 2d ago First seen · 425 lines · 21 tokens per session scan A 026eabe33287
do is a command published in the GitHub repository alphabravo-oss/guild (2 stars, last pushed 2d ago), licensed MIT. It adds 21 tokens to every session and 3,849 once invoked, about $0.0001 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
claude-flow-swarm
Coordinate multi-agent swarms for complex tasks.
bootstrap
PACT session-start ritual — identify the session team (platform-provisioned), secretary spawn, paused-state surface, bootstrap marker.
ox-session-review
Command "ox-session-review" from sageox/ox, covering failure-mode watch-list (read first), from the ledger root. should print 0, phase 1 — scan & score (read-only), quality buckets (first match wins) and removal candidates.
add-agent
引导新增一个 Agent 适配器。用法 /add-agent.
handoff
Package current state for the next agent.
xdo
Direct development by the Main Agent.