Examples

The examples/ module contains small, runnable harnesses. Each isolates one pattern; each has its own README; most accept -scripted and run a deterministic fake client so you can watch the machinery without an API key.

Example	What it demonstrates
healthcheck	A world-verifying goal: the agent probes a fake server farm until every endpoint reports healthy. Schema + fan-out guardrails policing the calls.
releasegate	Pre/post guardrails around a workflow: the agent may only publish after every required check is green, and the goal confirms the publish actually happened.
hnupvote	Browser automation under pursue: a real chromedp session where the oracle is verified world state and a login wall forces the re-drive path.
spawn_worker	Hierarchical decomposition: a coordinator launches researcher / reviewer / coder tasks with agent_spawn, continues independently, and collects results explicitly under work-mode gates and a depth cap.
stuck_recovery	The decompose guardrail’s graduated response — pass, advisory, fatal — as an agent repeats a failing search, then recovers by delegating.
long_conversation	Compactor integration: a pressure-gated compactor keeps a long exploration inside the context window without orphaning tool calls.
shared_infra	Shared infrastructure without an LLM: retrieval indexing/search, workflow graph execution, checkpoints, HITL review, and JSONL traces.

Running them

# deterministic, no LLM
go run ./examples/healthcheck -scripted

# shared infra tour, no LLM or external services
go run ./examples/shared_infra

# real provider — flags are uniform across examples
go run ./examples/healthcheck -provider anthropic -model claude-sonnet-4-6
go run ./examples/releasegate -provider llamacpp

Provider selection goes through the same backends registry as everything else: -provider names a builtin, keys come from the standard env vars, -base-url (or LLAMACPP_BASE_URL etc.) overrides endpoints.

Reading order

Start with shared_infra for the non-LLM infrastructure pieces, then healthcheck — it’s the smallest complete loop with a real goal. Then releasegate for guardrails as policy, and stuck_recovery for what graduated advisories look like in practice. hnupvote is the one to read when you want to see pursue driving something genuinely flaky (a browser) to a verified outcome.

Each example’s README walks through its harness construction; examples/patterns.md in the repo collects the recurring shapes — scripted clients, fake filesystems, oracle design — you’ll reuse in your own harnesses.