Add self-improvement flywheel problem document - #43
Closed
ralphbean wants to merge 2 commits into
Closed
Conversation
Member
Author
|
Hold on this one. Itneeds to have its language repositioned on top of #43. |
Introduces a new problem doc exploring how fullsend learns from its own outputs and from human corrections to those outputs. This is distinct from production-feedback.md (which covers platform signals feeding into agent work) — the flywheel addresses how the quality of fullsend's own decisions feeds back into improving the system itself. Covers: the capture problem, how improvements propagate (instruction refinement, context enrichment, golden-set expansion), the self-proposal mechanism, bootstrapping sequence, and feedback asymmetry. Assisted-by: OpenCode claude-opus-4-6@default
…ction - Remove Konflux-specific references per the org-agnostic repositioning (PR #42, now merged) - Add "Deterministic tool extraction" subsection incorporating Marta's review insight from PR #38: agents discover which checks should exist, those get codified into linters/scanners/policies, and the non-deterministic scope shrinks over time — reducing AI coupling - Add cross-references between flywheel and operational observability docs - Add deterministic tool proposals to the self-proposal list Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
ralphbean
force-pushed
the
agent-self-improvement-flywheel
branch
from
March 24, 2026 18:56
539e8dc to
a8c5a4b
Compare
Member
Author
|
The org-agnostic repositioning is done (rebased on main which includes PR #42). Also incorporated Marta's review feedback from PR #38 — her insight about agents discovering which checks should become deterministic tools is now a full subsection ("Deterministic tool extraction") under "How improvements propagate." The two docs cross-reference each other. Ready for review again. |
Member
Author
|
Dropping this as its not exactly a problem in the way that the other problem documents are "problems". It's (really) more of an opportunity/feature. |
2 tasks
ralphbean
added a commit
that referenced
this pull request
May 11, 2026
Add lychee as a pre-commit hook and standalone `make lint-md-links` target to catch broken in-repo links and anchors in markdown files. Fix all existing broken links: - Wrong anchor fragment in ADR 0002 (heading was renamed) - Missing "repo-" in ADR 0013 filename (2 references) - Dangling self-improvement-flywheel.md links → point to closed PR #43 - Dangling experiments/guardrails-eval links → point to fullsend-ai/experiments repo - Wrong relative paths for ADR 0019 from web/ and web/admin/ - Wrong relative path for ADR 0023 from skills/writing-user-docs/ - Dangling design doc link in a plan checklist Noticed the need for this in PR #770. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
4 tasks
This was referenced Jul 8, 2026
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.
Introduces a new problem doc exploring how fullsend learns from its own outputs and from human corrections to those outputs. This is distinct from production-feedback.md (which covers platform signals feeding into agent work) — the flywheel addresses how the quality of fullsend's own decisions feeds back into improving the system itself.
Covers: the capture problem, how improvements propagate (instruction refinement, context enrichment, golden-set expansion), the self-proposal mechanism, bootstrapping sequence, and feedback asymmetry.
Assisted-by: OpenCode claude-opus-4-6@default