Skip to content

feat(skills): i-have-adhd — ADHD-friendly output shaping (port of ayghri/i-have-adhd, 38.9k★ MIT) - #107947

Open
teknium1 wants to merge 1 commit into
mainfrom
tools-scout/i-have-adhd-skill
Open

teknium1 wants to merge 1 commit into
mainfrom
tools-scout/i-have-adhd-skill

Conversation

@teknium1

Copy link
Copy Markdown
Collaborator

Adds i-have-adhd to the optional-skills catalog (optional-skills/communication/i-have-adhd) — an output-style skill that shapes responses so an ADHD reader can act on them: lead with the next action, number multi-step work, restate state each turn, specific time estimates, visible wins, no preamble/recap/closers. Session-persistent until "stop adhd mode".

Source and traction

  • Port of ayghri/i-have-adhd — MIT, 38.9k stars, created May 2026, actively pushed (Sep 10). Front page of HN this week (535 points: "A skill to stop coding agents from burying the answer"); repeatedly recommended in agent-skill roundups.
  • Upstream LICENSE vendored as LICENSE.txt; author credited in frontmatter and commit.

What changed vs upstream

  • Adapted to Hermes conventions: hardline frontmatter (≤60-char description, tags, platforms), added When to Use and Verification sections, harness references point at Hermes (todo_list), removed upstream's disable-model-invocation (Hermes optional skills are explicit installs; discovery via description is desired).
  • Folded in 4 clarity fixes surfaced by the cold-subagent live test: first-turn semantics for "restate state", advice-only handling for "make wins visible", estimate-audience note, rule 1 vs rule 8 precedence for diagnostics.

Validation

Check Result
scripts/run_tests.sh tests/skills/test_authoring_standards.py green
scripts/run_tests.sh tests/skills (full dir, 44 files) green
Cold-subagent live test (read skill → answer a cron/ModuleNotFoundError prompt under its rules → self-audit) SHIP, 10/10 rules pass; 4 friction findings folded in
Docs catalog row + sidebar + auto-generated per-skill page (generate-skill-docs.py); regen debris (agent-merge-conflict-arbiter stray page, path-separator churn) excluded
Competitor-name grep on all added surfaces clean

Trust / cost notes

Prose-only skill: zero dependencies, zero credentials, zero network access. Loaded only when installed (hermes skills install official/communication/i-have-adhd) and triggered.

Fit

No overlap with existing skills: one-three-one-rule (decision briefs) and humanizer/sepia (de-AI prose) solve different problems; this is response shape for accessibility. Placed in communication/ beside one-three-one-rule.

Infographic

i-have-adhd skill

…hri/i-have-adhd, 38.9k★ MIT)

Port of https://github.com/ayghri/i-have-adhd (MIT, Ayoub Ghriss) into the
optional-skills catalog under communication/. Output-style skill: lead with
the next action, number multi-step work, restate state each turn, specific
time estimates, visible wins, no preamble/recap/closers. Session-persistent
until the user says "stop adhd mode".

Why: 38.9k stars, front page of HN this week (535 pts), repeatedly
recommended across agent-skill communities; accessibility win with zero
dependencies and zero credentials. Adapted to Hermes conventions (frontmatter,
When to Use, Verification, todo_list reference) rather than copied verbatim;
upstream LICENSE vendored. Cold-subagent live test passed 10/10 rules; its
four friction findings (first-turn state, advice-only turns, estimate
audience, rule 1 vs 8 precedence) are folded into the skill text.

Docs: catalog row, sidebar entry, and the auto-generated per-skill page via
website/scripts/generate-skill-docs.py.
@github-actions

github-actions Bot commented Sep 11, 2026 •

Copy link
Copy Markdown

૮ >ﻌ< ა ci review

ran on d9a1074 — feat(skills): i-have-adhd — ADHD-friendly output shaping (po

❌ Job failures

Python tests / Run tests · View job

Job Python tests / Run tests failed.


⚠️ Warnings

CI timings · View report · View job

Wall time 5m19s vs 1m11s (+349.3%). 3 faster, 1 unchanged.

  • Detect affected areas: -48.0s
  • OSV scan / Emit review status: -9.0s
  • OSV scan / Scan lockfiles / osv-scan: -6.0s

OSV vulnerability scan · View job

80 known vulnerabilities found in pinned dependencies.

How to fix:

Review the findings in the Security tab. Update the affected dependencies if a patched version is available.

@alt-glitch alt-glitch added type/feature New feature or request P3 Low — cosmetic, nice to have tool/skills Skills system (list, view, manage) labels Sep 11, 2026
@teknium1

Copy link
Copy Markdown
Collaborator Author

The Python tests failure here is a main-tip breakage (5 tests, red on main push run 34550073898 too), unrelated to this skill-only diff. Fix: #107962 — once that lands this PR needs a rebase/rerun.

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

P3 Low — cosmetic, nice to have tool/skills Skills system (list, view, manage) type/feature New feature or request

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants