feat: reduce ambient context token usage across 4 workflows#46684
Merged
Conversation
8 tasks
…kflows - daily-cache-strategy-analyzer.md: add gh-proxy + cli-proxy tools, consolidate Phase 2 sub-sections 2.1/2.2/2.3 into single detection block (~854 char reduction) - changeset.md: add github.mode: gh-proxy to complete proxy coverage - pr-sous-chef.md: remove redundant rules 10-11 from token efficiency rules, trim Required nudges section by removing duplicate shell examples (~3031 char reduction) - agent-persona-explorer.md: deduplicate Phase 3 batch/per-scenario assessment templates (~905 char reduction) - Recompile all 258 workflows (0 errors) Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>
Copilot
AI
changed the title
[WIP] Optimize daily ambient context workflows
feat: reduce ambient context token usage across 4 workflows
Jul 19, 2026
Contributor
🤖 PR Triage
Workflow token optimization (removes redundant instructions, adds proxy config). Draft. Low urgency; can be batched with similar workflow-chore PRs.
|
This was referenced Jul 20, 2026
pelikhan
approved these changes
Jul 20, 2026
pelikhan
marked this pull request as ready for review
July 20, 2026 04:14
Collaborator
|
@copilot resolve the merge conflicts on this branch. |
Contributor
There was a problem hiding this comment.
Pull request overview
Reduces ambient prompt context across four agentic workflows while adding proxy declarations for GitHub access.
Changes:
- Consolidates repeated workflow instructions and examples.
- Adds
gh-proxyconfiguration. - Regenerates workflow lock files.
Show a summary per file
| File | Description |
|---|---|
.github/workflows/pr-sous-chef.md |
Condenses nudge and safe-output guidance. |
.github/workflows/pr-sous-chef.lock.yml |
Updates generated body metadata. |
.github/workflows/daily-cache-strategy-analyzer.md |
Consolidates cache detection and declares proxies. |
.github/workflows/daily-cache-strategy-analyzer.lock.yml |
Updates generated metadata hashes. |
.github/workflows/changeset.md |
Declares GitHub proxy mode. |
.github/workflows/changeset.lock.yml |
Regenerates CLI-proxy runtime configuration. |
.github/workflows/agent-persona-explorer.md |
Deduplicates assessment criteria. |
.github/workflows/agent-persona-explorer.lock.yml |
Updates generated body metadata. |
Review details
Tip
Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
- Files reviewed: 8/8 changed files
- Comments generated: 4
- Review effort level: Medium
| safeoutputs update_pull_request --pull_request_number 12345 --update_branch true --body "<!-- pr-sous-chef branch update -->" --operation append | ||
| ``` | ||
| - If `mergeStateStatus` is `CONFLICTING`, **skip this step entirely**. | ||
| - Otherwise, attempt `update_pull_request` with `update_branch: true` and a minimal append body marker including `pr-sous-chef` and the run URL. |
| - Uses a volatile cache key with `${{ github.run_id }}` **and no restore-key fallbacks** (guaranteed cold start every run) | ||
| > **Note — "last one wins" pattern**: `<prefix>-${{ github.run_id }}` + restore key `<prefix>-` is valid. Only flag `run_id` keys as misconfigured when there are **no restore-key fallbacks**. | ||
| - Never writes to `/tmp/gh-aw/cache-memory/` (tool does nothing useful) | ||
| - Reads stale data without checking timestamps |
Comment on lines
+45
to
+47
| cli-proxy: true | ||
| github: | ||
| mode: gh-proxy |
Comment on lines
+46
to
+47
| github: | ||
| mode: gh-proxy |
…xt-daily-optimizer # Conflicts: # .github/workflows/agent-persona-explorer.lock.yml # .github/workflows/changeset.lock.yml # .github/workflows/daily-cache-strategy-analyzer.lock.yml # .github/workflows/pr-sous-chef.lock.yml Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>
Contributor
Author
Resolved in d2111ee. All four conflicts were in auto-generated |
Copilot AI
added a commit
that referenced
this pull request
Jul 20, 2026
…ontext optimization (#46684) Co-authored-by: gh-aw-bot <259018956+gh-aw-bot@users.noreply.github.com>
Contributor
|
🎉 This pull request is included in a new release. Release: |
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.
Daily Ambient Context Optimizer flagged high structural fragmentation and missing proxy config across four workflows, with median first-request size of 22K chars and P95 at 24K.
Changes
daily-cache-strategy-analyzer.mdgh-proxy+cli-proxytotools:— both were absent, leaving GitHub API and CLI calls on the slower unauthenticated path## Phase 2: Detect Cache Issuesblock; removes duplicate bash loop and table template (~854 chars saved)changeset.mdgithub: { mode: gh-proxy }—cli-proxy: truewas set but GitHub API calls were unoptimizedpr-sous-chef.md## Token efficiency rules— both restate the system-level safe-output terminal guarantee already injected at runtime## Required nudgesshell examples: consolidate the twoadd_commentexamples into one block, remove the standaloneupdate_pull_request/resolve_pull_request_review_thread/dismiss_pull_request_reviewexamples (format is established in the remaining block); all constraints preserved (~3,031 chars saved)agent-persona-explorer.md{{#if batch}}/{{else}}branches — both branches were repeating the identical 5-question checklist and quality-score structure (~905 chars saved)Total reduction: ~4,790 chars across 4 workflows. All 258 workflows recompile cleanly (0 errors).