Conversation
…tion Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>
|
@copilot review uses of aw_info.json in agent job and ensure it is using needs.activation.outputs... |
…oad activation artifact Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>
Fixed in fae505a. The failing Warning Firewall rules blocked me from connecting to one or more addresses (expand for details)I tried to connect to the following addresses, but was blocked by firewall rules:
If you need me to access, download, or install something from one of these locations, you can either:
|
|
💫 TO BE CONTINUED... Smoke Claude failed! Our hero faces unexpected challenges... |
|
📰 BREAKING: Smoke Copilot is now investigating this pull request. Sources say the story is developing... |
|
🌑 The shadows whisper... Smoke Codex failed. The oracle requires further meditation... |
|
@copilot move generate workflow overview to the activation job |
|
📰 DEVELOPING STORY: Smoke Copilot reports failed. Our correspondents are investigating the incident... |
Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>
Done in f8f1217. The "Generate workflow overview" step is now in the activation job, placed after Warning Firewall rules blocked me from connecting to one or more addresses (expand for details)I tried to connect to the following addresses, but was blocked by firewall rules:
If you need me to access, download, or install something from one of these locations, you can either:
|
|
📰 BREAKING: Smoke Copilot is now investigating this pull request. Sources say the story is developing... |
|
🌑 The shadows whisper... Smoke Codex failed. The oracle requires further meditation... |
|
🎬 THE END — Smoke Claude MISSION: ACCOMPLISHED! The hero saves the day! ✨ |
|
@copilot make sure audit command can find aw_info.json and prompt.txt in https://github.com/github/gh-aw/actions/runs/22524436360/job/65254212703 |
Agent Container Tool Check
Result: 12/12 tools available ✅ Overall Status: PASS
|
|
🤖 Smoke Test Report — §22524720391
Overall:
|
There was a problem hiding this comment.
Smoke test review of PR #18843 — clean refactor moving generate_aw_info to the activation job. The consolidation of aw_info.json + prompt.txt into a single activation artifact is a nice improvement. One minor nit on a misleading verbose log message in flattenActivationArtifact; otherwise the implementation looks solid. The backward compatibility path for old prompt artifacts is a thoughtful touch.
📰 BREAKING: Report filed by Smoke Copilot
| } | ||
| logsDownloadLog.Printf("Moved file: %s → %s", path, destPath) | ||
| if verbose { | ||
| fmt.Fprintln(os.Stderr, console.FormatVerboseMessage(fmt.Sprintf("Flattened: %s → %s", relPath, relPath))) |
There was a problem hiding this comment.
Minor nit: the verbose message logs relPath → relPath (same value for both source and destination), which makes the flattening output misleading. Consider logging the actual source and destination paths:
fmt.Fprintln(os.Stderr, console.FormatVerboseMessage(fmt.Sprintf("Flattened: activation/%s → %s", relPath, relPath)))| c.generateCreateAwInfo(&awInfoYaml, data, engine) | ||
| steps = append(steps, awInfoYaml.String()) | ||
| // Expose the model output from the activation job so downstream jobs can reference it | ||
| outputs["model"] = "${{ steps.generate_aw_info.outputs.model }}" |
There was a problem hiding this comment.
✅ Clean approach exposing model as an activation job output — this makes the model value available earlier in the pipeline and avoids duplication across 162+ lock files. The $\{\{ needs.activation.outputs.model }} reference in the agent job is much cleaner than the old $\{\{ steps.generate_aw_info.outputs.model }} pattern.
|
Commit pushed:
|
|
📰 VERDICT: Smoke Copilot has concluded. All systems operational. This is a developing story. 🎤 |
…json and prompt.txt Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>
Added
The |
|
Commit pushed:
|
Smoke Test Results - Run 22524720389
Overall: PARTIAL — Serena
|
There was a problem hiding this comment.
💥 Automated smoke test review - all systems nominal! The activation job refactoring looks solid — moving generate_aw_info earlier ensures the model context is available during prompt generation, which is the right architectural decision.
💥 [THE END] — Illustrated by Smoke Claude
|
@copilot update compiler: "generate agent info" step should be the first step in activation, after setup of action |
Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>
Done in b34d8e0. Warning Firewall rules blocked me from connecting to one or more addresses (expand for details)I tried to connect to the following addresses, but was blocked by firewall rules:
If you need me to access, download, or install something from one of these locations, you can either:
|
aw_info.jsonwas generated in the agent job, too late for it to be bundled with the prompt and available at activation time. This movesgenerate_aw_infointo the activation job and merges the workflow overview summary directly into that step, consolidating all activation files into a singleactivationartifact.Compiler changes
compiler_activation_jobs.go: Runsgenerate_aw_infoas the first step after setup in the activation job (beforevalidate-secret,validate-context-variables, and all other steps); exposesmodelas an activation job output; renames the uploaded artifact fromprompttoactivationand includes bothaw_info.jsonandprompt.txtcompiler_yaml.go: Adds a step-levelenv:section togenerate_aw_infoso the model env var resolves fromvars.*in the activation job context; merges the workflow overview summary call (generateWorkflowOverview) into the end of thegenerate_aw_infoinline script — there is now a single combined step; removes the standalonegenerateWorkflowOverviewStepfunctioncompiler_yaml_main_job.go: Removes thegenerate_aw_infostep; removes thegenerate_workflow_overviewstep; downloads theactivationartifact (instead ofprompt) to/tmp/gh-awto restore both files; removesaw_info.jsonfromagent-artifactspaths; updatesmodeloutput to${{ needs.activation.outputs.model }}Activation job step order (new)
Setup Scripts— sets up/opt/gh-aw/actionsGenerate agentic run info— first step after setup; writes/tmp/gh-aw/aw_info.json, sets model output, and writes the step summary viagenerateWorkflowOverview.githuband.agentsfoldersUpload activation artifact— uploadsaw_info.json+prompt.txtActivation artifact structure (new)
CLI logs/audit
logs_download.go: AddsflattenActivationArtifactto handle the multi-fileactivation/artifact downloaded bygh run download, movingaw_info.jsonandaw-prompts/prompt.txtto the expected root-level paths. Old single-filepromptartifacts remain handled byflattenSingleFileArtifactsfor backward compatibility.logs_download_test.go: AddsTestFlattenActivationArtifactto verify that after flattening,aw_info.jsonlands at the root output directory (required by the audit command for engine detection) andaw-prompts/prompt.txtlands at its expected path.All 162 workflow lock files recompiled; golden tests updated.
Original prompt
🔒 GitHub Advanced Security automatically protects Copilot coding agent pull requests. You can protect all pull requests by enabling Advanced Security for your repositories. Learn more about Advanced Security.
Changeset
activationartifact so the prompt generator and logs/audit can readaw_info.jsonearlier.Warning
The following domain was blocked by the firewall during workflow execution:
github.comTo allow these domains, add them to the
network.allowedlist in your workflow frontmatter:See Network Configuration for more information.
✨ PR Review Safe Output Test - Run 22524436360
Changeset
aw_info.jsonalong with the prompt in the newactivationartifact so activation-time flows and logs read the data sooner.Warning
The following domains were blocked by the firewall during workflow execution:
codeload.github.comgithub.comTo allow these domains, add them to the
network.allowedlist in your workflow frontmatter:See Network Configuration for more information.
✨ PR Review Safe Output Test - Run 22524720389