Pipecat: offline framework-adapter simulation

Twin: examples/sdk_framework_adapter_pipecat_process_promotion.py · emits agent-learning.run.v1 · offline, no credentials. A coding agent can complete this page from the frontmatter alone.

1. What you are testing

Pipecat coverage in the kit is probe-promoted, and this page is deliberately offline: it tests the frame-pipeline adapter, not a live audio transport. The twin, examples/sdk_framework_adapter_pipecat_process_promotion.py, builds a local LocalPipecatPipeline with two surfaces: a run(input) path that returns content with no frame trace and no tool evidence, and a process(payload) path that returns verified evidence only when the adapter passes metadata.framework == "pipecat" in the payload dict. Promotion selects the frame entrypoint with evidence; the text path is recorded as weak.

The failure class this catches is pipeline-shape mismatch: a voice harness that drives the convenience text surface can look healthy while the frame-processing contract a real pipeline uses goes untested. Pinning the adapter shape offline means the behavioral contract is already proven before any live transport enters the picture.

The run manifest, examples/framework_pipecat_manifest.json, drives the same adapter from the CLI. It targets the factory framework_shims.py:build_pipecat_pipeline with trace_runtime: true and replays a framework_trace environment whose span is pipeline.process. Everything runs on the local_text engine in one turn: offline, deterministic, no provider keys.

2. Run it

CLI (the required_env key is CI metadata for this offline manifest — any placeholder value satisfies it):

AGENT_LEARNING_MULTI_FRAMEWORK_EXAMPLE_KEY=local-example \
agent-learn run examples/framework_pipecat_manifest.json \
  --output artifacts/framework-pipecat.json

SDK, same operation (export the same placeholder env first):

import asyncio
from fi.alk import simulate

result = asyncio.run(
    simulate.run_manifest_file("examples/framework_pipecat_manifest.json")
)
assert result["kind"] == "agent-learning.run.v1"

3. What you built

Postcondition (machine-checkable — the same check the docs gate pattern uses):

python -c "import json; p=json.load(open('artifacts/framework-pipecat.json')); assert p['kind']=='agent-learning.run.v1', p['kind']; print('ok')"

The artifact carries status, a per-turn transcript for the framework-owner scenario, the evaluation report, and the framework runtime trace evidence the adapter extracted — including the pipeline.process span the manifest replays. It is a replayable record, not a log line: the same file feeds baseline, compare, and replay.

4. When it fails

Symptom	First-mile class	Doctor check
vendored import failed	infra	agent-learn doctor → summary.missing_engine_modules
missing required environment variable(s)	config fault	set the placeholder key shown above; agent-learn doctor → summary.public_boundary_passed confirms the install surface
transcript shows the weak run(input) path (no frame trace, no tool calls)	behavior regression	re-run the twin promotion and compare process evidence against the text fallback

5. Prove it / keep it

The twin is admitted by the framework_adapter_probe_readiness release gate, so every agent-learn release-check re-executes this exact frame-pipeline promotion path — the page stays true or the release fails. To keep your own pipeline honest, promote the run artifact into a regression baseline with the baseline / promote-to-regression / compare command family, then wire the manifest into CI. Live Pipecat transports (real audio in and out) are an opt-in lane (see ROADMAP.md: voice lane rungs) — this page stays on the offline golden path. The reader's job here is maintenance of a living proof, not a one-off demo.