Skip to content

feat: incremental planning and verified completion - #253

Merged
jkyberneees merged 4 commits into
mainfrom
feat/evidence-based-completion
Sep 18, 2026
Merged

jkyberneees merged 4 commits into
mainfrom
feat/evidence-based-completion

Conversation

@jkyberneees

@jkyberneees jkyberneees commented Sep 18, 2026

Copy link
Copy Markdown
Contributor

Odek can now revise plans during a turn while preserving completed work and acceptance checks.

  • Add, edit, reorder, split, or replace steps, with revision reasons saved across resume.
  • Require successful tool evidence for checked steps; replanning cannot silently drop requirements. Missing verification is flagged in the final answer.
  • Add 11 deterministic runtime eval scenarios and update the planning and developer documentation.

Planning remains model-directed. Checks verify tool success, not whether the chosen checks fully prove task correctness.

Validation: CLI tests, affected-package race tests, all 11 eval scenarios, lint, vulnerability checks, and documentation/schema checks passed locally.

@cloudflare-workers-and-pages

cloudflare-workers-and-pages Bot commented Sep 18, 2026

Copy link
Copy Markdown

Deploying with  Cloudflare Workers  Cloudflare Workers

The latest updates on your project. Learn more about integrating Git with Workers.

Status Name Latest Commit Preview URL Updated (UTC)
✅ Deployment successful!
View logs
odek b549d29 Commit Preview URL

Branch Preview URL
Sep 18 2026, 07:41 AM

Comment thread internal/eval/eval.go Fixed
Comment thread internal/eval/eval.go Fixed
@jkyberneees jkyberneees changed the title feat: evidence-based plan completion and runtime evals feat: incremental planning and verified completion Sep 18, 2026
@jkyberneees
jkyberneees merged commit 4b79cff into main Sep 18, 2026
10 checks passed
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.

2 participants