Agent Lifecycle

Last Updated: 2026-06-11

Primary Workflow Names

The core lifecycle should use primary task-verb tools. Each is implemented by a raw tool with the same parameters and identity rules, and returns a normalized envelope: the operationally useful fields first (next_action, state_summary, risk_summary, memory_suggestions, recovery_hint), with the full raw payload preserved under raw_governance.

Use the primary workflow tools by default. Use raw implementation names only for older servers, compatibility code, or when you explicitly need the unwrapped handler response.

Starting a Session

Choose creation, lineage, or proof-owned resume explicitly:

~~~text start_session(force_new=true) # any fresh session — the default; co-location is not lineage start_session(force_new=true, parent_agent_id="<dispatcher-uuid>", spawn_reason="subagent") # dispatched subagent (usually set automatically by the dispatcher) start_session(force_new=true, parent_agent_id="<exited-uuid>", spawn_reason="explicit") # handoff from an EXITED prior session identity(agent_uuid="<uuid>", continuity_token="<token>", resume=true) # same live owner / proof-owned rebind ~~~

Declaring a currently-live agent as parent is rejected (lineage_coincidental_rejected): a live agent is a concurrent sibling, not a predecessor. subagent and compaction are exempt — their parent is legitimately live. A genuine handoff to an exited predecessor stays provisional until R1 confirms it.

Use raw onboard(...) instead when targeting older servers or when you need the unwrapped raw response.

Returns:

• agent_uuid / UUID: The server identity anchor for this process instance
• client_session_id: In-session transport continuity metadata
• continuity_token: Short-lived ownership proof for PATH 0 anti-hijack, not indefinite cross-process continuity
• session diagnostics: session_resolution_source, identity_assurance, and deprecation warnings when relevant

Creation, lineage, and resume (updated 2026-04-25)

name= is a cosmetic label, not a resume key. Passing the same name on a later session does not prove identity.

Default rules:

1. Any fresh session: call start_session(force_new=true) with no parent. Co-location in a workspace is not lineage.
2. Declare lineage only for a real causal event — a dispatched subagent (parent_agent_id="<dispatcher-uuid>", spawn_reason="subagent", usually set automatically by the dispatcher) or a handoff from an EXITED prior session (parent_agent_id="<exited-uuid>", spawn_reason="explicit"). Declaring a currently-live agent as parent is rejected.
3. Same live process or explicit ownership rebind: call identity(agent_uuid="<uuid>", continuity_token="<token>", resume=true).
4. Ordinary check-ins: rely on the active session binding or client_session_id; reserve continuity_token for explicit proof-owned rebinds.

Avoid these patterns:

• Bare identity(agent_uuid=X, resume=true): UUID alone is an unsigned claim. It currently logs/emits hijack-suspected telemetry and is strict-mode rejected when UNITARES_IDENTITY_STRICT=strict.
• onboard(continuity_token=...) as cross-process resume: S1-a accepts this only during the deprecation window and returns a warning. Declare lineage with parent_agent_id instead.
• Bare onboard(): older code may still pin-resume by weak session/IP:UA evidence. Use force_new=true when creating a new process identity.

continuity_token is now intentionally narrow: 1-hour TTL, rolling, and retained as possession proof for anti-hijack gates. It does not establish process-instance continuity by itself.

Check-ins

Call sync_state() after meaningful work:

~~~text sync_state( response_text: "Brief summary of what you did", complexity: 0.0-1.0, # Task difficulty estimate confidence: 0.0-1.0 # How confident you are (be honest) ) ~~~

Use raw process_agent_update(...) when you need the raw handler payload; primary workflow responses preserve it under raw_governance.

When to Check In

• After completing a meaningful unit of work
• Before and after high-complexity tasks
• When you feel uncertain or notice drift
• Not after every single tool call — use judgment between these bounds

What You Get Back

A verdict plus current EISV metrics. Read the verdict and act on it.

Reading Verdicts

A guide verdict is an early warning. Ignoring it makes pause more likely.

Identity

• UUID is an identity anchor, not proof that the current process owns that identity
• Session binding can happen via transport session, client_session_id, or short-lived continuity token
• Use identity() when continuity seems unclear
• Inspect:
• identity_status
• bound_identity
• session_resolution_source
• continuity_token_supported
• identity_assurance
• deprecations

Strong ownership proof is better than implicit continuity. If the runtime falls back to weak signals such as fingerprinting, mint a fresh process identity and declare lineage.

Recovery

When you are paused, stuck, or need intervention:

Recovery is not a shortcut — self_recovery() examines your EISV state and determines if resumption is safe. If your metrics are genuinely degraded, it will not force a resume.

MCP Tools Reference

Essential (use in every session)

• start_session(force_new=true, parent_agent_id=...) — Create a fresh process identity, optionally declaring lineage
• sync_state() — Check in with work summary, complexity, confidence
• check_working_state() — Read your current EISV state
• identity() — Confirm who the runtime thinks you are and how continuity was resolved; include continuity_token for proof-owned UUID rebinds
• health_check() — Check operator-facing server health when behavior seems odd
• search_shared_memory(query=...) — Find existing knowledge before creating new entries
• knowledge(action="note", ...) — Quick contribution to the knowledge graph

Common (use when needed)

• knowledge() — Full knowledge graph CRUD (store, update, details, cleanup)
• agent() — Agent lifecycle (list, archive, get details)
• calibration() — Check or update calibration data
• request_dialectic_review() — Start a dialectic session
• export() — Export session history

Specialized

• call_model() — Delegate to a secondary LLM for analysis
• detect_stuck_agents() — Find unresponsive agents
• self_recovery() — Resume from stuck or paused state
• submit_thesis() / submit_antithesis() / submit_synthesis() — Dialectic participation