Issues

Chat GPT

ayghri/i-have-adhd#271 · opened

on Oct 29, 2026

1

Third-party AI-assisted report on i-have-adhd

ayghri/i-have-adhd#270 · opened

on Oct 26, 2026

Rule proposal: put anything the reader copies in a code block

ayghri/i-have-adhd#267 · opened

on Oct 21, 2026

always-on hook keeps an empty frontmatter block in the injected ruleset (Node and OpenCode paths; sh/PowerShell strip it)

ayghri/i-have-adhd#263 · opened

on Oct 18, 2026

1

Always-on hook misses the flag file when `CLAUDE_CODE_SUBPROCESS_ENV_SCRUB` removes `CLAUDE_CONFIG_DIR`

ayghri/i-have-adhd#255 · opened

on Oct 12, 2026

Add .devin plugin

ayghri/i-have-adhd#249 · opened

on Oct 5, 2026

3

Decode Pi smoke-test output as UTF-8 on Windows

ayghri/i-have-adhd#245 · opened

on Sep 29, 2026

Reject invalid eval scores, trials, and budgets

ayghri/i-have-adhd#243 · opened

on Sep 29, 2026

Feature / Preset Suggestion: Executive & Decision-Driven preset for complex engineering tasks

ayghri/i-have-adhd#241 · opened

on Sep 26, 2026

Multi-Agent Response Standard: Structured Output for Parallel Agents

ayghri/i-have-adhd#232 · opened

on Sep 20, 2026

SessionStart always-on hook can't run without Node, but Node is no longer a Claude Code dependency

ayghri/i-have-adhd#221 · opened

on Sep 16, 2026

Release gate can never pass: absolute no-blockers rule × the unpassable agent-owned-edit case

ayghri/i-have-adhd#218 · opened

on Sep 16, 2026

Copilot install docs list unsupported personal skill directory

ayghri/i-have-adhd#217 · opened

on Sep 16, 2026

evals: add RTL language coverage

ayghri/i-have-adhd#197 · opened

on Sep 11, 2026

Proposal: unified, reversible multi-agent installation

ayghri/i-have-adhd#193 · opened

on Sep 11, 2026

A lot of overhead, couldn't this be an output-style instead of a skill?

ayghri/i-have-adhd#187 · opened

on Sep 11, 2026

5

--allow-unmetered turns off the budget check

ayghri/i-have-adhd#181 · opened

on Sep 10, 2026

1

Clarify expected output length in README examples

ayghri/i-have-adhd#178 · opened

on Sep 10, 2026

hooks: PowerShell always-on fallback keeps trailing blank lines, so it disagrees with the Node and sh hooks

ayghri/i-have-adhd#172 · opened

on Sep 10, 2026

SessionStart hook output is plain text; Codex rejects it as invalid JSON and drops the ruleset

ayghri/i-have-adhd#156 · opened

on Sep 10, 2026

Performance optimization analysis: reduce unnecessary I/O, context overhead, and evaluation runtime

ayghri/i-have-adhd#153 · opened

on Sep 9, 2026

1

Non-object judge responses stop the entire batch

ayghri/i-have-adhd#151 · opened

on Sep 9, 2026

Add i-have-adhd to awesome-ai-plugins?

ayghri/i-have-adhd#141 · opened

on Sep 5, 2026

5

Proposal: deliver the ruleset to subagents and forks, ship it as a Claude Code output style, make always-on a plugin option

ayghri/i-have-adhd#137 · opened

on Sep 4, 2026

1

SessionStart always-on hook silently no-ops when CLAUDE_PLUGIN_ROOT is not exported (Claude Code desktop app, macOS)

ayghri/i-have-adhd#129 · opened

on Aug 25, 2026

Navigation Menu

Repository navigation

Issues

Chat GPT

Third-party AI-assisted report on i-have-adhd

Rule proposal: put anything the reader copies in a code block

always-on hook keeps an empty frontmatter block in the injected ruleset (Node and OpenCode paths; sh/PowerShell strip it)

Always-on hook misses the flag file when `CLAUDE_CODE_SUBPROCESS_ENV_SCRUB` removes `CLAUDE_CONFIG_DIR`

Add .devin plugin

Decode Pi smoke-test output as UTF-8 on Windows

Reject invalid eval scores, trials, and budgets

Feature / Preset Suggestion: Executive & Decision-Driven preset for complex engineering tasks

Multi-Agent Response Standard: Structured Output for Parallel Agents

SessionStart always-on hook can't run without Node, but Node is no longer a Claude Code dependency

Release gate can never pass: absolute no-blockers rule × the unpassable agent-owned-edit case

Copilot install docs list unsupported personal skill directory

evals: add RTL language coverage

Proposal: unified, reversible multi-agent installation

A lot of overhead, couldn't this be an output-style instead of a skill?

--allow-unmetered turns off the budget check

Clarify expected output length in README examples

hooks: PowerShell always-on fallback keeps trailing blank lines, so it disagrees with the Node and sh hooks

SessionStart hook output is plain text; Codex rejects it as invalid JSON and drops the ruleset

Performance optimization analysis: reduce unnecessary I/O, context overhead, and evaluation runtime

Non-object judge responses stop the entire batch

Add i-have-adhd to awesome-ai-plugins?

Proposal: deliver the ruleset to subagents and forks, ship it as a Claude Code output style, make always-on a plugin option

SessionStart always-on hook silently no-ops when CLAUDE_PLUGIN_ROOT is not exported (Claude Code desktop app, macOS)