ewilliams28/openclaw
1
0
Fork 0
mirror of https://github.com/openclaw/openclaw.git synced 2026-08-26 04:15:48 -06:00
Code Issues Packages Projects Releases Wiki Activity

fix(qa): track progress card planning signal (#129143)

Browse Source 
Amp-Thread-ID: https://ampcode.com/threads/T-01a037b4-f621-7144-bfdf-a68df795dd57

Co-authored-by: Amp <amp@ampcode.com>
This commit is contained in:
Peter Steinberger
2026-08-25 02:30:56 -07:00
committed by GitHub
parent fda20a6ed2
commit daefcc117f
2 changed files with 8 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files
extensions/qa-lab/src/scenario-flow-runner.test.ts
+3 -3
View File
@@ -231,10 +231,10 @@ function createPlanningEvidenceFixture(
return {
scenario,
outboundText: `Built ${artifactFile}`,
failureMessage: "missing OpenClaw update_plan signal",
failureMessage: "missing OpenClaw progress_card signal",
currentSummary: {
eventCursor: 9,
successfulToolCallCounts: { update_plan: 1 },
successfulToolCallCounts: { progress_card: 1 },
},
};
}
@@ -254,7 +254,7 @@ function runPlanningEvidenceFixture(
{ identity: "old-turn:plan", text: "Codex plan:\nQA_INTERNAL_PLAN_DO_NOT_SEND" },
{ identity: "old-turn:assistant", text: fixture.outboundText },
],
successfulToolCallCounts: { update_plan: 1 },
successfulToolCallCounts: { progress_card: 1 },
},
currentSummary,
];
qa/scenarios/workspace/medium-game-plan-openclaw-harness.yaml
+5 -5
View File
@@ -12,7 +12,7 @@ scenario:
successCriteria:
- A live-frontier run fails fast unless the selected primary model is openai/gpt-5.6-luna.
- The scenario forces the embedded OpenClaw harness before the build turn.
- The OpenClaw harness records a successful `update_plan` tool call before the artifact assertion passes.
- The OpenClaw harness records a successful `progress_card` tool call before the artifact assertion passes.
- The agent writes a self-contained HTML game with a canvas loop, controls, scoring, waves, pause, and restart.
docsRefs:
- docs/plugins/sdk-agent-harness.md
@@ -35,7 +35,7 @@ scenario:
gameTitle: Star Garden Defenders
minBytes: 5000
buildPrompt: |-
Call update_plan with a short implementation plan before editing.
Call progress_card with a short implementation plan before editing.
Then build a medium-complex, self-contained browser game at ./star-garden-defenders-openclaw.html.
@@ -108,8 +108,8 @@ flow:
- afterEventCursor:
expr: transcriptBefore.eventCursor
- assert:
expr: "(transcript.successfulToolCallCounts.update_plan ?? 0) > 0"
message: missing OpenClaw update_plan signal
expr: "(transcript.successfulToolCallCounts.progress_card ?? 0) > 0"
message: missing OpenClaw progress_card signal
- set: artifactPath
value:
expr: "path.join(env.gateway.workspaceDir, config.artifactFile)"
@@ -150,4 +150,4 @@ flow:
expr: "outbound.text.includes(config.artifactFile)"
message:
expr: "`final reply did not mention ${config.artifactFile}: ${outbound.text}`"
detailsExpr: "`provider=${selected?.provider} model=${selected?.model} runtime=${config.harnessRuntime} agentRun=${turn.started.runId} appServerTurn=n/a planSignal=update_plan:success:${transcript.successfulToolCallCounts.update_plan ?? 0} artifact=${artifactPath} bytes=${artifact.length} visibleReply=${JSON.stringify(outbound.text)}`"
detailsExpr: "`provider=${selected?.provider} model=${selected?.model} runtime=${config.harnessRuntime} agentRun=${turn.started.runId} appServerTurn=n/a planSignal=progress_card:success:${transcript.successfulToolCallCounts.progress_card ?? 0} artifact=${artifactPath} bytes=${artifact.length} visibleReply=${JSON.stringify(outbound.text)}`"