Skills issues on macOS — page 2
Issues the maintainers labelled both platform:macos and area:skills.
477 issues · 82 open · 395 resolved (83%) · first seen Feb 23, 2026
Is this getting better or worse?
This class of problem is converging. 163 new reports in the last 90 days vs 286 in the 90 before — -43%. The open backlog peaked at 117 in 2026-04 and sits at 82 today.
Derived by counting each matching issue from its creation date until it closed. First-party
computation over the anthropics/claude-code tracker — the trend is not published
anywhere upstream.
Which builds report this
316 of these reports name the Claude Code build they were running, spanning 116 releases. Heaviest reporters:
- Claude Code v2.1.119
- Claude Code v2.1.87
- Claude Code v2.1.112
- Claude Code v2.1.63
- Claude Code v2.1.207
- Claude Code v2.1.92
- Claude Code v2.1.76
- Claude Code v2.1.233
Report volume, not a defect count: an issue can mention a build without that build having caused the failure. Versions come from the version string in the bug-report template, matched against real published tags.
Do these get fixed?
Median time to close is 32 days across 395 closures. Of the 395 closures with a recorded reason, 24% were closed as completed and 299 as not-planned or duplicate. The most recent completed fix landed in 2026-08.
Workarounds reported by the community
113 of these issues post a workaround someone says worked. The highest-engagement ones:
- [Bug] /loop slash command not recognized in v2.1.71
- …d on a recurring interval (e.g. /loop 5m check the deploy)" Workaround The underlying CronCreate tool works correctly and achieves the same functionality: - CronCreate with cron expression /5 and a prompt successfully sc…
Found in the description of #31759 · resolved - AskUserQuestion auto-submits without waiting for user input in skill/command context
- …t was never collected until they review the final artifact. Workaround implemented Added defensive instructions to skill workflow prompts: after each AskUserQuestion call, verify the result contains actual option labels.…
Found in the thread of #29674 · resolved - [BUG] Skill `model:` frontmatter triggers 429 "Extra usage required" on 2.1.76 when session uses opus[1m]
- …Sonnet promoted to . Rejected if account lacks entitlement. Workarounds - Stay on 2.1.75 — skill model switching works correctly without auto-promotion. Caveat: if a sonnet skill (200K) is invoked from an opus[1m] sessio…
Found in the description of #34296 · still open - [BUG] permissions.additionalDirectories does not load skills from added directories (--add-dir CLI flag does)
- …ilable-skills list. Only bundled skills ( , , etc.) appear. Workaround: Restored — all 22 skills load fine via the personal-scope path. Confirming this is identical symptom to the Windows repro in 43267 (auto-closed as d…
Found in the thread of #30064 · resolved - Regression: built-in `/code-review` rejects invocation from another skill (disable-model-invocation), breaking workflow composition
- …estore this without changing the safer default. Our current workaround is to pause the workflow and hand the developer an explicit manual step, then resume with the output folded in — but that reintroduces exactly the ma…
Found in the thread of #79560 · resolved
Excerpts are quoted from the linked issue so you can judge relevance before opening it. Threads that only ask for a workaround, or report that there is none, are excluded. These are community suggestions, not official guidance from Anthropic — read the full thread before applying one.
Matching issues 101–200
Ranked by community engagement (reactions weighted over comments). Listing the top 300 of 477 matches; the statistics above cover all 477.
- [BUG] /schedule skill doesn't surface custom MCP connectors for routine attachment
- /insights report only analyzes 328/1291 sessions, missing most data
- [BUG] /ultrareview blocks on "enable overage billing" while /extra-usage reports unlimited extra usage is already enabled
- VS Code extension: skill invocation no longer shows visual badge (regression since ~2.1.110)
- [Bug] Skill invocation and instruction retention degradation
- [BUG] Plugin skill tab completion: short name completes but qualified name is not discoverable
- Skills broken in 2.1.88, work correctly in 2.1.87
- No way to enforce approval on all file modifications
- [BUG] Skill loads into context and works via the internal Skill tool, but /skill-name and /plugin:skill slash commands return "Unknown skill"
- Project-level .claude/commands/ files not loaded into skills list
- [BUG] doctor is behaving oddly - no longer instant
- [BUG] Classifier blocks user-authorized actions inside forked skills — parent-turn intent is invisible to the fork, and no per-action override exists
- [BUG] Skill tool re-invoked after slash command already loaded skill
- Built-in deep-research workflow: Scope phase exhausts StructuredOutput retry cap on long briefs
- Desktop app Code tab does not discover plugin commands (only skills)
- [BUG] Desktop app missing slash commands available in CLI (/btw, /compact, /diff, etc.)
- [BUG] --add-dir CLI flag does not load skills from added directory's .claude/skills/ (v2.1.89)
- Plugin skills not discovered when .claude/ directory exists in plugin
- Project-level skills not taking precedence over plugin marketplace skills
- claude -p does not trigger skills — skill-creator eval loop always shows 0% recall
- Skills with model: field fail with 'Rate limit reached' when model differs from main session
- Project-level skills in .claude/skills/ not discovered
- Claude Code repeatedly ignores CLAUDE.md rules — manually generates artifacts instead of using configured skills
- Skill deny rules in settings.json do not block user-initiated slash commands
- [BUG] command autocomplete only triggers at position 0, while @ triggers mid-prompt
- Nested code-review fan-out spawned 877 agents despite 5-level depth limit
- [BUG] Skill name conflict persists after full local cleanup — registry appears server-side
- Typing `//` no longer filters autocomplete to project skills whose frontmatter `name:` starts with `/` (regression in v2.1.160)
- [BUG] Excessive token consumption in web search workflows - exponential token usage growth
- [Bug] org-plugins skills unavailable until 10-minute periodic sync fires — race condition between SkillsPlugin sync and custom-3p:org-plugins install on every startup
- [BUG] `/insights` skill: AI-generated sections empty (returns {})
- [BUG] Claude Desktop 1.8089.1: marketplace migration race makes user-installed skills unreachable via slash-command
- Skills in subdirectory .claude/skills/ not discovered in monorepo
- [Bug] Skill description truncation warning lacks clarity on dropped skills and target context allocation
- [Bug] Project-local skill override not loaded by runtime
- [BUG] [ultrareview] Free runs consumed on large-scope branch reviews with empty findings (2 sessions)
- Claude Code conducts symptom-driven exploration instead of root-cause analysis on complex multi-component systems — concrete side-by-side with Codex showing the gap
- Regression: AskUserQuestion still unavailable in context: fork skills (closed #46654 reoccurring)
- [BUG] [ultrareview] Run crashed server-side, consumed free credit (1 of 3) with zero findings produced
- [Bug] /ultrareview misleading behavior: reviews only branch diffs, not full codebase
- Skills edited via Customize UI not reflected in Code tab (storage path mismatch)
- Model drifts from mandatory process rules during long batch sessions
- [BUG] ~/.claude/skills/ directory deleted on first run after 2.1.113→2.1.114 update (native binary migration)[BUG]
- [BUG] `/ultrareview` misses untracked files — sends `git diff` only, no client-side warning
- [Feature Request] Allow customizing skill trigger character to separate from built-in commands (e.g. `$` instead of `/`)
- [Bug] Missing /buddy Command - Command Not Found
- Custom slash commands that chain other slash commands drop remaining steps after the first nested invocation
- [BUG] /buddy command completley dissapered a soon as i tried making a plug in for it
- Cowork: skills visible in Customize but not loaded in chat (context budget)
- Skill-scoped hooks in SKILL.md frontmatter are not passed to forked subagent when using context: fork
- Custom skill in .claude/skills/ recognized but fails with 'Unknown skill' when invoked
- [BUG] When Bash(mkdir:*) is listed in allowed-tools, Claude still prompts for permission
- Skill `model:` frontmatter ignored in interactive sessions (works in `claude -p`) — v2.1.228
- [BUG] Plugin/Skill "Update" button stays disabled even when a newer version exists upstream
- Built-in dynamic workflows (deep-research) have no model routing — all sub-agents inherit the main-loop model, burning Opus on grunt work
- [BUG] Skill tool broken in headless (`claude -p`) mode since v2.1.141 — returns is_error, skill body never loaded
- [BUG] ultrareview crashes before producing findings — credits consumed but no results returned
- [BUG] Max user, Claude unusable 3x issues - 1. 1M context error 2. repeated auto compaction after 1/2 messages 3. Single word prompt "too long"
- commit-commands plugin uses session working directory instead of the git repo where changes were made
- [BUG] /ultrareview burned free quota on server-side rate-limit failure
- ultrareview crashed before producing findings — refund request
- [BUG] Report: Claude Code repeatedly violates loaded workflow skill, escalates hallucination across correction rounds
- [BUG] /ultrareview failed TWICE
- [BUG] Plugin install does not seed userConfig.*.default into settings.json — surfaces as skill !-context bash failing "Contains expansion"
- [BUG] editing my type to censor literal text "ultraplan"
- [Bug] Project-level skills not included in system prompt
- /ultrareview consistently failing with upstream service errors (502, 520, socket close)
- [Bug] Rate limit error on /ultrareview consumes free attempt quota
- [BUG] /ultrareview bug - Twice for the Review fail in operation : Second run crashed again with the same error — no findings delivered.
- [BUG] /ultrareview Find step fails with empty [] for local repos — 3 attempts, all failed, and consumed all 3 free credits
- [BUG] Ultra-Review Failing
- Ultrareview crash still counts against free usage limit
- [BUG] .skill writes from channels still does not work with DSP
- [BUG] Stale plugin installs from abandoned git worktrees silently break TUI slash-command autocomplete
- [BUG] bypassPermissions mode still prompts for .claude/skills edits despite documented exemption
- [BUG] Bring Loamwise back — /buddy
- [BUG] /ultrareview only checks uncommitted files, not full branch history
- [Bug] Permission bypass not working for .claude/skill directory creation
- [BUG] Getting Please run /login · API Error: 403 on skill load when using with LiteLLM
- [BUG] LLM is trying to invoke a non-exist tool: \"Skill\"
- Marketplace plugin source.path not used as root for skill discovery
- Third-party marketplace plugins cannot register skills/commands/agents
- [BUG] Unknown skill when CLAUDE_CONFIG_DIR is set and skill lives in $CLAUDE_CONFIG_DIR/skills/
- [BUG] User-level skills from ~/.claude/skills/ appear in Desktop Code tab autocomplete but don't invoke
- [BUG] Skill triggers ignored even with non-huge context
- Slash commands for plugin skills return 'Unknown skill' with custom CLAUDE_CONFIG_DIR
- /voice command triggers plugin skills instead of activating voice mode
- [Bug] Ralph-loop: Backticks in quoted prompts passed to Bash causing command substitution error
- [BUG] Claude Code no longer load skills invoked by slash commands
- Project skills silently dropped beyond ~28 limit — no warning or error
- Skill tool aborts with "Sibling tool call errored" when a parallel tool call fails
- Plugin --plugin-dir namespacing breaks skill allowed-tools resolution
- [BUG] Skill invocations persist across /clear and context compaction
- code-review skill: one nested background finder agent's task-notification silently dropped; parent has no fallback and returns fabricated status instead of its required output
- [Bug] `/code-review` exhausts entire usage allowance due to runaway recursive behavior
- Skill frontmatter `model:` override ignored in interactive mode but honored in print mode (v2.1.227)
- /reload-skills does not refresh the slash-command autocomplete menu
- Worktree isolation inflates spawned-shell env past ARG_MAX → E2BIG on all Bash, unrecoverable without restart
- /dataviz skill listed in release notes but not available in session
- [BUG] Post-compaction: a no-argument "mode" skill is re-read as active after the user's in-chat disable is dropped from the summary