Skip to content

fix(researcher): always emit a safe output when no opportunities found#16

Merged
JackTreble merged 2 commits into
mainfrom
copilot/debug-workflow-failure
Mar 25, 2026
Merged

fix(researcher): always emit a safe output when no opportunities found#16
JackTreble merged 2 commits into
mainfrom
copilot/debug-workflow-failure

Conversation

Copy link
Copy Markdown
Contributor

Copilot AI commented Mar 25, 2026

The Market Research Discovery workflow could complete successfully while producing zero safe outputs — triggering a false-positive failure issue — because the prompt instructed the agent to "create 0 issues" when no candidates survived validation but never required it to call noop.

Changes

  • Step 3: Annotated the "0 issues" rule with an explicit forward reference requiring noop to be called
  • New Step 6 — Always produce a safe output: Formalizes the invariant:
    • 1+ issues created → create-issue calls already satisfy the requirement
    • 0 issues created → agent must call noop with a brief explanation of what was reviewed and why each candidate was rejected
Original prompt

This section details on the original issue you should resolve

<issue_title>[aw] Market Research Discovery failed</issue_title>
<issue_description>### Workflow Failure

Workflow: Market Research Discovery
Branch: main
Run: https://github.com/JackTreble/Web-Tools/actions/runs/23567512336

⚠️ No Safe Outputs Generated: The agent job succeeded but did not produce any safe outputs.

This typically indicates:

  • The safe output server failed to run
  • The prompt failed to generate any meaningful result
  • The agent should have called noop to explicitly indicate no action was taken

Action Required

Option 1: Debug using any coding agent

Use this prompt with any coding agent (GitHub Copilot, Claude, Gemini, etc.):

Debug the agentic workflow failure using https://raw.githubusercontent.com/github/gh-aw/main/debug.md

The failed workflow run is at https://github.com/JackTreble/Web-Tools/actions/runs/23567512336

Option 2: Assign this issue to Copilot

Assign this issue to Copilot using the agentic-workflows sub-agent to automatically debug and fix the workflow failure.

Option 3: Manually invoke the agent

Debug this workflow failure using your favorite Agent CLI and the agentic-workflows prompt.

[!TIP]
To stop a workflow from creating failure issues, set report-failure-as-issue: false in its frontmatter:

safe-outputs:
  report-failure-as-issue: false

Generated from Market Research Discovery ·

  • expires on Apr 1, 2026, 10:41 PM UTC

Comments on the Issue (you are @copilot in this section)


💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.

…issues created

Co-authored-by: JackTreble <4872989+JackTreble@users.noreply.github.com>
Agent-Logs-Url: https://github.com/JackTreble/Web-Tools/sessions/ec7fe7ff-1eff-4d0b-a100-c666577191db
Copilot AI changed the title [WIP] Debug agentic workflow failure for market research fix(researcher): always emit a safe output when no opportunities found Mar 25, 2026
Copilot AI requested a review from JackTreble March 25, 2026 22:53
@JackTreble JackTreble marked this pull request as ready for review March 25, 2026 22:57
@JackTreble JackTreble merged commit abe12b3 into main Mar 25, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

[aw] Market Research Discovery failed

2 participants