dld-goal: execution loop, and Stage 1 complete - #33
Closed
jimutt wants to merge 2 commits into
Closed
Conversation
Completes Stage 1 of the goal loop: - /dld-goal continue works one item as a four-part transaction (claim, mechanical check, review, record), checking bounds and drift first - failure handling follows DL-004: one retry with the failure as context, then block and ask the operator; skip leaves decisions proposed and says so in the report - start now runs guard-preconditions before creating anything, and resume re-validates before any work - documents that the skill deliberately holds no loop machinery, and that manual pacing is a real workflow rather than a degraded one Packaging: README skills table and long-running-features section, the steering rule, dld-init's CLAUDE.md block, and the run contract doc. Suite at 305. Co-Authored-By: Claude Opus 5 <noreply@pi.dev> Generated-By: pi 0.84.2
Co-Authored-By: Claude Opus 5 <noreply@pi.dev> Generated-By: pi 0.84.2
jimutt
force-pushed
the
feat/dld-goal-skill
branch
from
August 21, 2026 11:49
a6978e2 to
f48b7d0
Compare
Owner
Author
|
Superseded by feat/pi-harness-extension — all commits from this branch are merged there. See #41 for the findings log on top. |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Stacks on #32. Completes Stage 1 and accepts DL-003, DL-004, DL-005.
What
/dld-goal continueworks one item as a four-part transaction: claim, mechanical check, review subagent, record. A model saying "done" never completes an item.Failure handling follows DL-004 — one retry with the failure as context, then block and ask.
skipleaves decisionsproposedand the run report says so, so a partial run can't read as a clean finish.Packaging: README, steering rule, dld-init's CLAUDE.md block.
Notable
The skill deliberately contains no loop machinery. Without the Pi extension a run is paced by the user — each
/dld-goal continueadvances one item, with.dld/runs/carrying everything between invocations. That's a real workflow, and it's stated plainly in the skill so it isn't oversold.Validated end to end in a scratch repo: batched items verify together, a missing annotation and a failing check both report, blocking is refused before the retry is used,
next-itemrefuses to step past a blocker, skipping leaves decisions proposed, and a decision edited while the run was paused blocks the resume.Suite at 305,
tessl plugin lintclean.