Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
26 commits
Select commit Hold shift + click to select a range
cf4a0dd
docs: design for unified harness tracing/message-emitting surface
declan-scale Jun 18, 2026
4538544
docs: refine unified harness spec — span derivation rules, TurnUsage,…
declan-scale Jun 18, 2026
4573d18
docs: link deferred tool-error decision to AGX1-371
declan-scale Jun 18, 2026
8eda56d
docs: foundation implementation plan for unified harness surface (PRs…
declan-scale Jun 18, 2026
de56483
feat(harness): foundation types for unified harness surface
declan-scale Jun 18, 2026
a13725c
test(harness): cover CloseSpan defaults and HarnessTurn runtime check
declan-scale Jun 18, 2026
13868ee
feat(harness): pure SpanDeriver reducing the canonical stream to span…
declan-scale Jun 18, 2026
0ecc03f
refactor(harness): deterministic flush order + defensive index/orphan…
declan-scale Jun 18, 2026
8d708f4
feat(harness): SpanTracer adapter from span signals to adk.tracing
declan-scale Jun 18, 2026
7955d55
refactor(harness): guarded make_logger import + lifecycle contract te…
declan-scale Jun 18, 2026
803191b
feat(harness): yield_events delivery adapter (passthrough + tracing)
declan-scale Jun 18, 2026
dab044f
refactor(harness): simplify yield_events guard + cover finally-flush …
declan-scale Jun 18, 2026
3cc0326
feat(harness): auto_send delivery adapter (canonical stream -> adk.st…
declan-scale Jun 18, 2026
260064e
refactor(harness): exception-safe full-message post + drop dead state…
declan-scale Jun 18, 2026
b27367b
feat(harness): UnifiedEmitter facade tying delivery + tracing + usage
declan-scale Jun 18, 2026
ed86a46
refactor(harness): inject streaming into UnifiedEmitter + cover auto_…
declan-scale Jun 18, 2026
b5f6b94
test(harness): conformance scaffold + CI integration job skeleton
declan-scale Jun 18, 2026
520849a
test(harness): match scripts/test invocation + document conformance r…
declan-scale Jun 18, 2026
a915170
refactor(harness): isinstance narrowing for clean type-check across t…
declan-scale Jun 18, 2026
e7b9c52
refactor(harness): narrow auto_send tracer guards, drop type:ignore f…
declan-scale Jun 18, 2026
ebc468d
style: ruff import-sort + format fixes across the harness package
declan-scale Jun 18, 2026
8b0da83
fix(harness): mark overridden start_span with @override for pyright (…
declan-scale Jun 18, 2026
f9266cf
fix(harness): relative import in conformance test for pyright (report…
declan-scale Jun 18, 2026
b538187
fix(harness): index-keyed routing, tool stream delivery, final_text l…
declan-scale Jun 18, 2026
dcd65b5
test(harness): add tests for AGX1-377 tool stream delivery, index rou…
declan-scale Jun 18, 2026
b4b8b33
feat(harness): thread created_at through UnifiedEmitter.auto_send_tur…
declan-scale Jun 18, 2026
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
40 changes: 40 additions & 0 deletions .github/workflows/harness-integration.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,40 @@
name: Harness Integration

on:
push:
branches: [main]
pull_request:
paths:
- "src/agentex/lib/core/harness/**"
- "src/agentex/lib/adk/_modules/**"
- ".github/workflows/harness-integration.yml"

jobs:
conformance:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2

- name: Install uv
uses: astral-sh/setup-uv@d4b2f3b6ecc6e67c4457f6d3e41ec42d3d0fcb86 # v5.4.2
with:
version: '0.10.2'

- name: Bootstrap
run: ./scripts/bootstrap

# Defer to scripts/test so the harness suite runs under the exact same
# invocation as the main CI test job: DEFER_PYDANTIC_BUILD=false and
# `uv run --isolated --all-packages --all-extras pytest`, across the
# min/max supported Python versions. Running `uv run pytest` directly
# would risk an all-extras-only dep passing locally but failing in CI.
- name: Conformance suite
run: ./scripts/test tests/lib/core/harness/ -v

# Live integration matrix (harness x {sync, async, temporal}) is added per-harness
# in the migration plans. Placeholder job keeps the workflow valid until then.
live-matrix:
runs-on: ubuntu-latest
if: false # enabled once the first harness's test agents land
steps:
- run: echo "populated by migration PRs" # TODO(harness-migration): enable per-harness; see docs/superpowers/plans migration PRs 4-8
Loading
Loading