Repository navigation
Gates refuse a sprint in the run-data dir; record run 24 (haiku-supervisor experiment) - #41
Merged
Conversation
…visor experiment) Live run 24 ran wordfreq with a haiku supervisor: 10/13, $0.66 in 51 turns (sonnet: $0.76 in 24), and the supervisor planned and built the whole sprint under .aladdin/sprint1/, the git-ignored run-data dir. Every gate passed it. gate-stage2-complete and gate-dispatch now refuse a sprint or STORY_DIR inside .aladdin/ or .agentic/ and name docs/agents/sprintN/; with no argument the stage-2 gate looks in docs/agents/ first. Tests written first (red, then green). make e2e-live takes MODEL= for the supervisor so the experiment reproduces as a make target. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_014j4wLin3oJTtnvjE9Lq2VV
adeelahmad
pushed a commit
that referenced
this pull request
Oct 18, 2026
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_015QxyVf3NgwsgRyYpBgDDTC
adeelahmad
pushed a commit
that referenced
this pull request
Oct 18, 2026
…not .aladdin/ SKILL.md's layout tree listed `sprintN/` below `<project>/.aladdin/` (run data), and live run 24's haiku supervisor put the whole sprint in `.aladdin/sprint1/` (#41 made the gates refuse it). The tree now nests sprintN/ under docs/agents/ and draws .aladdin/ as run data only. scripts/test/skill-layout.test.py checks the tree and holds SKILL.md to a 50 KB budget so #43's trim does not grow back. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_015QxyVf3NgwsgRyYpBgDDTC
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.
Requested by Adeel · project thread
Before: a sprint could be planned and built anywhere. Live run 24, an experiment with a haiku supervisor on the two-story
wordfreqintent, put the whole sprint under.aladdin/sprint1/, the git-ignored run-data dir that is never committed. Every gate passed it, and the run's three planning checks found no sprint (10/13). The haiku supervisor also saved nothing: $0.66 in 51 turns against sonnet's $0.76 in 24 (run 23).After:
gate-stage2-completeandgate-dispatchrefuse a sprint dir or aSTORY_DIRinside.aladdin/or.agentic/, and the block namesdocs/agents/sprintN/and themvto get there. With no argument, the stage-2 gate looks indocs/agents/first.make e2e-livetakesMODEL=for the supervisor, so the experiment reproduces asmake e2e-live YES=1 INTENT=wordfreq MODEL=haiku. The supervisor stays on the strong route.How: a path-component check on the resolved sprint dir (stage-2) and on the contract's
STORY_DIR(dispatch), before the existing checks. ROADMAP M0.8 and the evidence ledger record run 24 as an experiment, not a public number; CHANGELOG, bin/README, SKILL.md and TRACKING are updated.Evidence: tests written first and red before the change:
planner-stuck.test.py6/6 (new: refused named or discovered, for both dirs) anddispatch.test.py13/13 (new: refused for both dirs, the docs/agents dir still passes).make test-hosts pycheck json conformance version-check shellcheck eval-validateOK,check-tracking46 requirements,claude plugin validate --strictpasses.Generated by Claude Code