v0.3.3

Highlights

• Busy-repository accuracy: the audit and optimizer workflows now fetch AI credit logs per workflow before aggregating results, preventing log truncation from skewing analysis when many workflows run concurrently in the same repository.
• Reliable optimizer targeting: the optimizer no longer treats completed runs with zero AI credits as optimization candidates, and no longer targets the audit and optimizer monitoring workflows themselves.

Generated by Release · 223 AIC · ⊞ 25.2K

v0.3.0

Highlights

• AI credits are now the primary billing metric. Both the audit and optimizer workflows track and report AI credit spend (AIC) aligned with the GitHub Copilot billing model introduced on June 1, 2026 (1 AIC = $0.01 USD). Audit issues and optimizer reports now lead with AI credit totals; effective token counts remain available as a legacy field.
• Audit trend chart now tracks active workflow volume alongside spend. The historical trend chart plots daily AI credit spend on the primary axis and the count of distinct workflows executed per day on the secondary axis, making spend-vs-activity correlation visible at a glance.

Upgrade Notes

• Chart asset filenames have changed. token_by_workflow.png → ai_credits_by_workflow.png; token_trend.png → ai_credits_trend.png. Update any automation, notifications, or documentation that references these filenames.

Generated by Release · ● 8.3M

v0.2.2

Highlights

• Workflow reports and documentation no longer show dollar-denominated cost estimates. Token usage is now the sole unit reported, removing figures that varied by model pricing and could be misleading.

Note: This project is pre-1.0. Interfaces may change between minor releases.

Generated by Release · ● 1.1M

v0.2.1

This release is primarily maintenance. No user-facing workflow behavior has changed.

Generated by Release · ● 2.5M

v0.2.0

Release notes are being prepared.

v0.1.1

🌟 Release Highlights

This release publishes the latest workflow sources and compiled lock files, updated since v0.1.0.

Included workflow sources:

• copilot-token-audit.md — Copilot token audit workflow
• copilot-token-optimizer.md — Copilot token optimizer workflow

Each workflow ships with a compiled .lock.yml file for pinned, reproducible execution.

Generated by Release · ● 772.4K

---

What's Changed

• Add release automation agentic workflow by @Copilot in #13
• Remove copilot-requests feature and compiled request permissions from token audit workflow by @Copilot in #14
• Upgrade to gh-aw v0.72.1 by @mnkiefer in #15
• chore: improve local testing of published workflows by @mnkiefer in #16
• fix: use latest version by @Copilot in #17
• ci: add git configuration by @mnkiefer in #18
• ci: add missing permissions by @mnkiefer in #19
• ci: install workflows for execution by @agentic-ops-bot[bot] in #20
• Fix audit workflow agent startup failure when OTLP secrets are unset by @Copilot in #23

New Contributors

• @mnkiefer made their first contribution in #15
• @agentic-ops-bot[bot] made their first contribution in #20

Full Changelog: v0.1.0...v0.1.1

v0.1.0

Added

• Initial published set of GitHub Agentic Workflows for Copilot token auditing and optimization.
• copilot-token-audit, which collects recent Copilot workflow usage and stores daily audit snapshots.
• copilot-token-optimizer, which analyzes high-token workflows and proposes conservative efficiency improvements.