bootstrap

A session-start procedure for PACT, a workflow that organizes coding work among specialized agents. It explains how to identify the session team, restore paused work, and use the project configuration.

In plain words
What is it for?
Use it at the start or resumption of a PACT coding session to identify the team, check saved state, and prepare specialist-agent coordination.
Why use it?
It gives agents a consistent way to begin or resume a session without creating duplicate teams or losing workflow state. The excerpt does not describe the full PACT process.

Command

Part of the pact-plugin plugin — 21 skills, 13 commands, 13 agents, 11 hooks shipped together

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/synaptic-labs-ai/pact-plugin/bootstrap
Clone the repo
git clone --depth 1 https://github.com/Synaptic-Labs-AI/PACT-Plugin

Or install pact-plugin, the plugin that ships this one along with the rest of its 21 skills, 13 commands, 13 agents, 11 hooks.

Per session 24 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,186 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 $0.00024 $0.02186
Opus 5 $0.00012 $0.01093
Sonnet 5 $0.00005 $0.00437
Haiku 4.5 $0.00002 $0.00219

Measured 3d ago against content hash 4a769f5c79d6, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

bootstrap 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 3d 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.

pact-plugin/commands/bootstrap.md · 80 lines

How it starts

The opening of the file, as written. The whole thing — 80 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Session-Start Ritual

The persona body's §2 Session-Start Ritual is your invocation contract; this command holds the mechanical detail. Execute the steps below in order, substituting Session Placeholder Variables from your context.


Step 1 — Identify the session team

Read team_name from the Current Session block in the project's CLAUDE.md (preferred location: $CLAUDE_PROJECT_DIR/.claude/CLAUDE.md; legacy fallback: $CLAUDE_PROJECT_DIR/CLAUDE.md). The session_init hook writes this block at session start.

The platform manages exactly one team per session, named {team_name} — it is provisioned automatically; you do not create it (the TeamCreate/TeamDelete tools no longer exist). Use {team_name} for every specialist dispatch.

Team-config expectations (bidirectional): on resume, a PRESENT team config is reused — never re-create it. An ABSENT config is the NORM after a clean session end — the platform provisions the session team; both states are normal, neither is corruption. Ghost-detection: if a spawn or send fails with a team-not-found error, do NOT blind-retry Agent() — the platform spawn is non-atomic and the process may already be running. Check for inbound messages from the "failed" agent name first; re-spawn only if it stays silent.

Step 2 — Spawn pact-secretary

Spawn the session secretary using single-task dispatch — the pact-secretary agentType is exempt from the teachback gate. No Task A teachback round-trip.

  1. TaskCreate(subject="secretary: deliver session briefing", description="<full mission: deliver session briefing on spawn, answer memory queries during the session, process HANDOFFs at workflow boundaries; CONTEXT / MISSION / INSTRUCTIONS / GUIDELINES per the orchestrator persona §13 Recommended Agent Prompting Structure>") — single work task. The subject names a discrete deliverable (the briefing), NOT the secretary's standing role; the standing duties (memory queries, HANDOFF harvest) live in the description as mission context and are tracked by their own later tasks.
  2. TaskUpdate(task_id, owner="secretary") — assign to the secretary; no addBlockedBy (no teachback gate)
  3. Agent(name="secretary", team_name="{team_name}", subagent_type="pact-secretary", description="Spawn session secretary", prompt="YOUR PACT ROLE: teammate (secretary).\n\nYou are joining team {team_name}. As your FIRST action, Invoke Skill(\"PACT:pact-team-registration\") to record your identity. Then check TaskList for tasks assigned to you.")
    • Use subagent_type="pact-secretary" and the canonical name="secretary" — the literal name is load-bearing.

Read the full file on GitHub · 80 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. 3d ago First seen · 80 lines · 24 tokens per session scan A 4a769f5c79d6

Subscribe to this mod's changes

bootstrap is a command published in the GitHub repository Synaptic-Labs-AI/PACT-Plugin (71 stars, last pushed 3d ago), licensed MIT. It adds 24 tokens to every session and 2,186 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 2026-08-30.