v2.3.0

What's Changed

• fix: F5 action self-test fixture uses the 3-row shape by @maumoya in #32
• feat: three STATUS modes, leaf, main and master (2.3.0) by @maumoya in #33

Full Changelog: v2.2.0...v2.3.0

v2.2.0

What's Changed

• feat: 3-row thread table, full grouped hub table (2.2.0) by @maumoya in #31

Full Changelog: v2.1.0...v2.2.0

v2.1.0

What's Changed

• feat: ETA on every open status row (2.1.0) by @maumoya in #30

Full Changelog: v2.0.0...v2.1.0

v2.0.0

What's Changed

• feat: thread deck, one pane for every open thread (v2.0.0) by @maumoya in #29

Full Changelog: v1.10.0...v2.0.0

v1.10.0

What's Changed

• fix: no repeat status tables on turns the user did not type (1.10.0) by @maumoya in #27
• fix: finished threads always end with the archive prompt (1.10.0) by @maumoya in #28

Full Changelog: v1.9.0...v1.10.0

v1.9.0

What's Changed

• feat: dynamic cards: numbers render from launch/facts.json by @maumoya in #24
• docs: demo video card at the top of the README by @maumoya in #25
• feat: all-done status prompts to archive and go back to the parent or roadmap thread (1.9.0) by @maumoya in #26

Full Changelog: v1.8.1...v1.9.0

v1.8.1

v1.8.1: the reply checker on the GitHub Marketplace

What changed in 1.8.1

• The action description is now under 125 characters, which the GitHub Marketplace requires, so the action is listed there as "ADHD status reply lint"
• No behavior change: same checks, annotations, job summary and problems output as 1.8.0

- uses: maumoya/adhd-status@v1.8.1
  with:
    files: |
      replies/**/*.md

Measured (1.7.0, 2026-11-03)

• Plugin eval, 11 cases, 3 runs each: 0.96 with the skill vs 0.50 without

Verify the zip
slsa-verifier verify-artifact adhd-status-v1.8.1.zip --provenance-path adhd-status-v1.8.1.intoto.jsonl --source-uri github.com/maumoya/adhd-status --source-tag v1.8.1

Full Changelog: v1.8.0...v1.8.1

v1.8.0

v1.8.0: the reply checker as a GitHub Action

What changed in 1.8.0

• The format checker now runs in CI: uses: maumoya/adhd-status@v1.8.0 with files: (paths or globs, one per line)
• Each problem shows as an annotation on the file, the job summary lists every file with its count, the problems output holds the total
• The step fails when a reply breaks the format, or when no file matches
• Listed on the GitHub Marketplace as "ADHD status reply lint"
• The skill itself is unchanged from 1.7.0

- uses: maumoya/adhd-status@v1.8.0
  with:
    files: |
      replies/**/*.md

Measured (1.7.0, 2026-11-03)

• Plugin eval, 11 cases, 3 runs each: 0.96 with the skill vs 0.50 without

Verify the zip
slsa-verifier verify-artifact adhd-status-v1.8.0.zip --provenance-path adhd-status-v1.8.0.intoto.jsonl --source-uri github.com/maumoya/adhd-status --source-tag v1.8.0

Full Changelog: v1.7.0...v1.8.0

v1.7.0

v1.7.0: never-miss status

What changed

• Loads on every way you say status: inside a sentence, after a compaction, and on every done or hand-back report
• Table first, plus one check before sending
• Hub mode: the hub only consolidates, and merge or deploy steps stay in the child thread
• Optional hooks make triggering certain (documented, never installed for you)

Measured (2026-11-03)

• Plugin eval (11 cases, 3 runs each): 0.96 with the skill vs 0.50 without
• Token bench vs 1.6.0 (4 episodes per arm): same spend per STATUS (median $0.084 vs $0.079); paste asks 0 vs 2
• Built from 86 real status turns, see docs/FAILURES.md

Verify the zip
slsa-verifier verify-artifact adhd-status-v1.7.0.zip --provenance-path adhd-status-v1.7.0.intoto.jsonl --source-uri github.com/maumoya/adhd-status --source-tag v1.7.0

v1.6.0

What's Changed

• docs: author photo in the README header by @maumoya in #3
• docs: show maumoya/adhd-status as the header title by @maumoya in #4
• docs: contributing guide + OpenSSF Best Practices badge by @maumoya in #5
• docs: about the author section by @maumoya in #6
• docs: skill publishing playbook by @maumoya in #7
• fix: working eval cases + skill proof rule (benchmark 0.39 → 0.95) by @maumoya in #8
• docs: public launch progress page by @maumoya in #9
• docs: measured benchmark results + re-prompt cost harness by @maumoya in #10
• docs: measured benefits at the top of the README by @maumoya in #11
• docs: big savings callout beside the results table by @maumoya in #12
• docs: result badges + language fix by @maumoya in #13
• bench: cheaper skill texts tested, all lose by @maumoya in #14
• Row 4 names every hidden item (1.5.0) by @maumoya in #15
• Look in logs before 'can't verify' (1.6.0) by @maumoya in #16
• Bench: no-paste rule lost; 1.6.0 stays by @maumoya in #17
• docs: README shows 1.5.0 and 1.6.0 results by @maumoya in #18

Full Changelog: v1.4.0...v1.6.0