Add Safe Output Guarantee to news-interpellations workflow#1646
Merged
Add Safe Output Guarantee to news-interpellations workflow#1646
Conversation
Contributor
🏷️ Automatic Labeling SummaryThis PR has been automatically labeled based on the files changed and PR metadata. Applied Labels: size-xs Label Categories
For more information, see |
Contributor
🔍 Lighthouse Performance Audit
📥 Download full Lighthouse report Budget Compliance: Performance budgets enforced via |
…news-interpellations workflow Fixes #1642 - The agent was running for 20 minutes but never calling a safe output tool (content length: 13 bytes, zero safe outputs). Added three critical prompt sections that were present in working news-propositions.md but missing from news-interpellations.md: 1. MANDATORY: Safe Output Guarantee section (top of prompt) 2. HARD DEADLINE time guard (minutes 43-45 in time budget) 3. FINAL REMINDER section (end of prompt) Since the lock file uses runtime-import, these changes take effect immediately without recompilation. Agent-Logs-Url: https://github.com/Hack23/riksdagsmonitor/sessions/8a789742-79b9-493a-8642-f042a09da4ef Co-authored-by: pethers <1726836+pethers@users.noreply.github.com>
github-actions
bot
added
documentation
Contributor
🔍 Lighthouse Performance Audit
📥 Download full Lighthouse report Budget Compliance: Performance budgets enforced via |
Copilot
AI
changed the title
[WIP] Fix workflow failure for News: Interpellation Debates
Add Safe Output Guarantee to news-interpellations workflow
Apr 10, 2026
Contributor
There was a problem hiding this comment.
Pull request overview
Updates the news-interpellations agentic workflow prompt to prevent “zero safe output” failures by adding the same guardrails already used in other news workflows.
Changes:
- Adds a top-of-prompt “Safe Output Guarantee” section with a 35-minute time guard.
- Adds a hard deadline step in the time budget to force a safe output before timeout.
- Adds a final end-of-prompt reminder requiring exactly one safe output call.
| - **Minutes 25–33**: Generate articles for core languages (EN, SV) using `npx tsx scripts/generate-news-enhanced.ts` | ||
| - **Minutes 33–38**: Validate and fix any quality issues | ||
| - **Minutes 38–43**: Commit analysis artifacts + articles, create PR with `safeoutputs___create_pull_request` | ||
| - **Minutes 43–45**: 🚨 **HARD DEADLINE** — If no safe output has been called yet, IMMEDIATELY call `safeoutputs___noop` with reason "Time limit reached before completion" |
There was a problem hiding this comment.
The HARD DEADLINE instruction unconditionally tells the agent to call safeoutputs___noop if no safe output has been called yet. This conflicts with the Safe Output Guarantee rule above (#1/#3), where the correct action near timeout is to create a PR if any artifacts exist. As written, the agent could generate/commit artifacts and then still be instructed to noop, dropping the work. Consider making the deadline conditional: create PR if there are any committed changes/artifacts; only noop when none exist (or MCP unreachable).
Suggested change
| - **Minutes 43–45**: 🚨 **HARD DEADLINE** — If no safe output has been called yet, IMMEDIATELY call `safeoutputs___noop` with reason "Time limit reached before completion" | |
| - **Minutes 43–45**: 🚨 **HARD DEADLINE** — If no safe output has been called yet, first check whether any analysis artifacts, generated articles, or committed changes exist. If any work exists, IMMEDIATELY call `safeoutputs___create_pull_request` to preserve it. Only call `safeoutputs___noop` with reason "Time limit reached before completion" when no artifacts or committed changes were produced, or when safe output/MCP is unavailable. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Agent ran 20 minutes, produced 13 bytes of output, zero safe output calls. The prompt lacked the three sections that force the agent to always call
safeoutputs___create_pull_requestorsafeoutputs___noopbefore session end. These sections were already present innews-propositions.mdbut missing fromnews-interpellations.md.Changes to
.github/workflows/news-interpellations.md:Minutes 43–45forces immediatenoopif no output yetLock file uses
{{#runtime-import}}, so no recompilation needed.