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/josix/agent-flow/team-orchestrategit clone --depth 1 https://github.com/josix/agent-flowWhat 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 | $0.00019 | $0.08070 |
| Opus 5 | $0.00010 | $0.04035 |
| Sonnet 5 | $0.00004 | $0.01614 |
| Haiku 4.5 | $0.00002 | $0.00807 |
Grade B, and why
team-orchestrate scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
MODE=$(grep '^mode:' .claude/team-orchestration.local.md | sed 's/mode: *//' | tr -d '"') How it starts
The opening of the file, as written. The whole thing — 887 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Team Orchestrate Command
Coordinate complex tasks through sequential delegation to specialist agents, with PARALLEL execution of Review + Verification phases using Agent Teams when available.
Arguments
--use-deep-dive: Use existing deep-dive context to skip or accelerate exploration phase--force-sequential: Force sequential mode even if Agent Teams is available<task description>: The task to orchestrate
State Initialization
FIRST: Initialize team orchestration state by running:
# Parse flags from arguments
USE_DEEP_DIVE=false
FORCE_SEQUENTIAL=false
TASK_ARGS="$ARGUMENTS"
if [[ "$ARGUMENTS" == *"--use-deep-dive"* ]]; then
USE_DEEP_DIVE=true
TASK_ARGS=$(echo "$TASK_ARGS" | sed 's/--use-deep-dive//' | xargs)
fi
if [[ "$ARGUMENTS" == *"--force-sequential"* ]]; then
FORCE_SEQUENTIAL=true
TASK_ARGS=$(echo "$TASK_ARGS" | sed 's/--force-sequential//' | xargs)
fi
# Build init command
INIT_CMD="bash ${CLAUDE_PLUGIN_ROOT}/scripts/init-team-orchestration.sh"
[[ "$USE_DEEP_DIVE" == true ]] && INIT_CMD="$INIT_CMD --use-deep-dive"
[[ "$FORCE_SEQUENTIAL" == true ]] && INIT_CMD="$INIT_CMD --force-sequential"
INIT_CMD="$INIT_CMD $TASK_ARGS"
eval "$INIT_CMD"
This creates .claude/team-orchestration.local.md to track:
- Current phase and iteration
- Orchestration mode (team vs sequential)
- Parallel group status (review_verification with sub-phases)
- Gate results for each phase
- Agent actions and timestamps
Prompt Refinement (Pre-Phase)
Before beginning orchestration, ensure the task is well-defined:
-
Check Task Clarity: Does "$ARGUMENTS" specify:
- What needs to be changed?
- Where in the codebase?
- What problem it solves?
-
If Vague: Ask ONE clarifying question before proceeding
- Provide options when possible
- Reference prompt-refinement skill for guidance
-
If Clear: Transform into structured format:
- Goal: One-sentence outcome
- Description: What and why (2-3 sentences)
- Actions: Concrete steps
- Constraints: Non-negotiable limits
- Assumptions: Things believed true that, if false, would change the approach
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 · 887 lines · 0 tokens per session scan B 2db9188c7673
team-orchestrate is a command published in the GitHub repository josix/agent-flow (7 stars, last pushed 17d ago), licensed MIT. It adds 19 tokens to every session and 8,070 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
riskreview
The user invoked the /riskreview facade command from the risk-review-pipeline pack.
music-suno-prompt
Grounded Suno prompt synthesis from local knowledge corpus + persona canon + label canon. No vibes-prompting.
lavra-work-teams
Work on multiple beads with persistent worker teammates that self-organize through a ready queue.
triage
Triage review findings one by one — present pending TODOs for user decision (approve, skip, modify).
lavra-import
Import a markdown plan into beads as an epic with child tasks.
audit-plugin
Audit plugin skills, commands, and agents for structure, size, and naming issues.