Repository navigation
fix(ci): port #1169 npm cache fix and run the harness from a snapshot - #197
Merged
Merged
Conversation
…and run the harness from a snapshot
… under the size guard
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Problem
Every skill run failed from about 2026-10-05 17:00 to 2026-10-06 19:00 UTC on both instances, for example tweet-digest https://github.com/aaronjmars/aeon-agent/actions/runs/37500275537 and fetch-tweets https://github.com/aaronjmars/miroshark-aeon/actions/runs/37500793087:
2.1.287 exists on npm. The "Cache Claude Code (npm)" step fell back through
restore-keysto an oldclaude-code-Linux-2.1.168cache. That cache's package index predates 2.1.287, and--prefer-offlinetrusts it, so the install fails. A failed job never saves the new cache key, so every run failed the same way until GitHub evicted the old cache. It has healed on its own (theclaude-code-Linux-2.1.287cache now exists and runs pass since 2026-10-06 19:00), but the next pin bump would break the same way. Canon already fixed this in aeonfun/aeon#1169. This ports it now instead of waiting for the next aeon-update sync.The same PR adds the run-harness snapshot from aeonfun/aeon#1180. aeon-update failed on 2026-10-05 (https://github.com/aaronjmars/aeon-agent/actions/runs/37300267512) because it rewrote
harness-adapter/run-harnesswhile that script was running. The workflow file is read from main when a run starts, so this has to be on the instance before the next aeon-update run.Changes
aeon.yml,messages.yml: droprestore-keyson the Claude Code npm cache (port of fix(ci): drop restore-keys on Claude Code npm cache so pin bumps install aeonfun/aeon#1169).aeon.yml+ newscripts/harness-adapter-snapshot.sh: runrun-harnessfrom a copy at$HOME/.aeon-run/harness-adapter, read-only inside the sandbox (same as fix(ci): run the harness from a snapshot so a skill that syncs harness-adapter cannot break its own run aeonfun/aeon#1180; in a script to keep the Run step under the 21000-character limit).Checks
scripts/tests/test_dry_run_dispatch.sh: 38 passed.