The agent loop is the serialized, per-session run that turns a message into actions and a reply: intake, context assembly, model inference, tool execution, streaming, persistence.

Entry points

• Gateway RPC: agent and agent.wait.
• CLI: openclaw agent.

Run sequence

1. agent RPC validates params, resolves the session (sessionKey/sessionId), persists session metadata, and returns { runId, acceptedAt } immediately.
2. agentCommand runs the turn: resolves model + thinking/verbose/trace defaults, loads the skills snapshot, calls runEmbeddedAgent, and emits a fallback lifecycle end/error if the embedded loop did not already emit one.
3. runEmbeddedAgent: serializes runs via per-session and global queues, resolves model + auth profile, builds the OpenClaw session, subscribes to runtime events, streams assistant/tool deltas, enforces the run timeout (aborting on expiry), and returns payloads plus usage metadata. For Codex app-server turns it also aborts an accepted turn that stops producing app-server progress before a terminal event.
4. subscribeEmbeddedAgentSession bridges runtime events to the agent stream: tool events to stream: "tool", assistant deltas to stream: "assistant", lifecycle events to stream: "lifecycle" (phase: "start" | "end" | "error").
5. agent.wait (waitForAgentRun) waits for lifecycle end/error on a runId and returns { status: ok|error|timeout, startedAt, endedAt, error? }.

Queueing and concurrency

Runs are serialized per session key (session lane) and optionally through a global lane, preventing tool/session races. Messaging channels choose a queue mode (steer/followup/collect/interrupt) that feeds this lane system; see Command Queue.

Transcript writes are additionally protected by a session write lock on the session file. The lock is process-aware and file-based, so it catches writers that bypass the in-process queue or come from another process. Writers wait up to session.writeLock.acquireTimeoutMs (default 60000 ms; env override OPENCLAW_SESSION_WRITE_LOCK_ACQUIRE_TIMEOUT_MS) before reporting the session as busy.

Session write locks are non-reentrant by default. A helper that intentionally nests acquisition of the same lock while preserving one logical writer must opt in with allowReentrant: true.

Session and workspace preparation

• Workspace is resolved and created; sandboxed runs may redirect to a sandbox workspace root.
• Skills are loaded (or reused from a snapshot) and injected into env and prompt.
• Bootstrap/context files are resolved and injected into the system prompt.
• A session write lock is acquired and the session transcript target is prepared before streaming starts. Any later transcript rewrite, compaction, or truncation path must take the same lock before mutating the SQLite transcript rows.

Prompt assembly

System prompt is built from OpenClaw's base prompt, skills prompt, bootstrap context, and per-run overrides. Model-specific limits and compaction reserve tokens are enforced. See System prompt for what the model sees.

Hooks

OpenClaw has two hook systems:

• Internal hooks (Gateway hooks): event-driven scripts for commands and lifecycle events.
• Plugin hooks: extension points inside the agent/tool lifecycle and gateway pipeline.

Internal hooks (Gateway hooks)

• agent:bootstrap: runs while building bootstrap files before the system prompt is finalized. Use it to add or remove bootstrap context files.
• Command hooks: /new, /reset, /stop, and other command events (see the Hooks doc).

See Hooks for setup and examples.

Plugin hooks

These run inside the agent loop or gateway pipeline:

Hook decision rules for outbound/tool guards:

• before_tool_call: { block: true } is terminal and stops lower-priority handlers. { block: false } is a no-op and does not clear a prior block.
• before_install: same terminal/no-op semantics as above. Use security.installPolicy, not before_install, for operator-owned install allow/block decisions that must cover CLI install and update paths.
• message_sending: { cancel: true } is terminal and stops lower-priority handlers. { cancel: false } is a no-op and does not clear a prior cancel.

See Plugin hooks for the hook API and registration details.

Harnesses can adapt these hooks. The Codex app-server harness keeps OpenClaw plugin hooks as the compatibility contract for documented mirrored surfaces; Codex native hooks are a separate, lower-level Codex mechanism.

Streaming

• Assistant deltas stream from the agent runtime as assistant events.
• Block streaming can emit partial replies on text_end or message_end.
• Reasoning streaming can be a separate stream or block replies.
• See Streaming for chunking and block reply behavior.

Tool execution

• Tool start/update/end events emit on the tool stream.
• Tool results are sanitized for size and image payloads before logging/emitting.
• Messaging tool sends are tracked to suppress duplicate assistant confirmations.

Reply shaping

Final payloads are assembled from assistant text (plus optional reasoning), inline tool summaries (when verbose and allowed), and assistant error text when the model errors.

• The exact silent token NO_REPLY is filtered from outgoing payloads.
• Messaging tool duplicates are removed from the final payload list.
• If no renderable payloads remain and a tool errored, a fallback tool error reply is emitted unless a messaging tool already sent a user-visible reply.

Compaction and retries

Auto-compaction emits compaction stream events and can trigger a retry. On retry, in-memory buffers and tool summaries reset to avoid duplicate output. See Compaction.

Event streams

• lifecycle: emitted by subscribeEmbeddedAgentSession (and as a fallback by agentCommand).
• assistant: streamed deltas from the agent runtime.
• tool: streamed tool events from the agent runtime.

The Gateway projects lifecycle and tool start/terminal events into the bounded, metadata-only audit ledger. This projection records provenance and result codes without copying prompts, messages, tool arguments, tool results, or raw errors out of the transcript/runtime path.

Chat channel handling

Assistant deltas buffer into chat delta messages. A chat final is emitted on lifecycle end/error.

Timeouts

Stuck session diagnostics

With diagnostics enabled, diagnostics.stuckSessionWarnMs (default 120000 ms) classifies long processing sessions with no observed reply, tool, status, block, or ACP progress:

• Active embedded runs, model calls, and tool calls report as session.long_running. Owned silent model calls stay session.long_running until diagnostics.stuckSessionAbortMs so slow or non-streaming providers are not flagged as stalled too early.
• Active work with no recent progress reports as session.stalled. Owned model calls switch to session.stalled at or after the abort threshold; ownerless stale model/tool activity is not hidden as long-running.
• session.stuck is reserved for recoverable stale session bookkeeping, including idle queued sessions with stale ownerless model/tool activity.

diagnostics.stuckSessionAbortMs defaults to at least 5 minutes and 3x the warn threshold. Stale session bookkeeping releases the affected session lane immediately after recovery gates pass; stalled embedded runs are abort-drained only after the abort threshold, so queued work resumes without cutting off merely slow runs. Recovery emits structured requested/completed outcomes; diagnostic state is marked idle only if the same processing generation is still current, and repeated session.stuck diagnostics back off while the session stays unchanged.

Where things can end early

• Agent timeout (abort)
• AbortSignal (cancel)
• Gateway disconnect or RPC timeout
• agent.wait timeout (wait-only, does not stop the agent)

Related

• Tools - available agent tools
• Hooks - event-driven scripts triggered by agent lifecycle events
• Compaction - how long conversations are summarized
• Exec Approvals - approval gates for shell commands
• Thinking - thinking/reasoning level configuration

Source: docs/concepts/agent-loop.md