gilles/claude-code-best-practice
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge remote-tracking branch 'origin/main'

Browse Source
This commit is contained in:
Shayan Rais
2026-05-12 23:51:39 +05:00
parent 0bea2799c9 1f001b277c
commit 2239d1ce04
4 changed files with 88 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
README.md
+3 -3
View File
@@ -157,8 +157,8 @@ Repos primarily known as curated libraries of subagent definitions (`.claude/age
| Name | ★ | <img src="!/tags/a.svg" height="14"> |
|------|---|---|
| [msitarzewski/agency-agents](https://github.com/msitarzewski/agency-agents) | 95k | 185 |
| [VoltAgent/awesome-claude-code-subagents](https://github.com/VoltAgent/awesome-claude-code-subagents) | 19k | 144 |
| [msitarzewski/agency-agents](https://github.com/msitarzewski/agency-agents) | 96k | 198 |
| [VoltAgent/awesome-claude-code-subagents](https://github.com/VoltAgent/awesome-claude-code-subagents) | 20k | 189 |
<p align="center">
<img src="!/claude-jumping.svg" alt="section divider" width="60" height="50">
@@ -361,7 +361,7 @@ Repos primarily known as curated libraries of subagent definitions (`.claude/age
## 🎬 VIDEOS / PODCASTS
| Video / Podcast | Source | YouTube |
|-----------------|--------|---------|
|-----------------|--------|--------|
| From Vibe Coding to Agentic Engineering (Andrej) \| 02 May 2026 \| AI Engineer | [![Karpathy](!/tags/community-karpathy.svg)](https://x.com/karpathy) | [YouTube](https://www.youtube.com/watch?v=96jN2OCOfLs) |
| Full Walkthrough: Workflow for AI Coding (Matt) \| 24 Apr 2026 \| Matt Pocock | [![Matt](!/tags/community-matt.svg)](https://x.com/mattpocockuk) | [YouTube](https://youtu.be/-QFHIoCo-Ko) |
| Everything We Got Wrong About Research-Plan-Implement (Dex) \| 24 Mar 2026 \| MLOps Community | [![Dex](!/tags/community-dex.svg)](https://x.com/daborhyde) | [YouTube](https://youtu.be/YwZR6tc7qYg) |
best-practice/claude-settings.md
+6
View File
@@ -94,6 +94,7 @@ Within the managed tier, precedence is: server-managed > MDM/OS-level policies >
| `disableDeepLinkRegistration` | string | - | Set to `"disable"` to prevent Claude Code from registering the `claude-cli://` protocol handler with the operating system on startup. Deep links let external tools open a Claude Code session with a pre-filled prompt via `claude-cli://open?q=...`. The `q` parameter supports multi-line prompts using URL-encoded newlines (`%0A`). Useful in environments where protocol handler registration is restricted or managed separately |
| `showThinkingSummaries` | boolean | `false` | Show extended thinking summaries in interactive sessions. When unset or `false` (default in interactive mode), thinking blocks are redacted by the API and shown as a collapsed stub. Redaction only changes what you see, not what the model generates — to reduce thinking spend, lower the budget or disable thinking instead. Non-interactive mode (`-p`) and SDK callers always receive summaries regardless of this setting |
| `disableSkillShellExecution` | boolean | `false` | Disable inline shell execution for `` !`...` `` and `` ```! `` blocks in skills and custom commands from user, project, plugin, or additional-directory sources. Commands are replaced with `[shell command execution disabled by policy]` instead of being run. Bundled and managed skills are not affected (v2.1.91) |
| `skillOverrides` | string | - | Control automatic skill invocation behavior. Values: `"off"` (skills are not invoked at all), `"user-invocable-only"` (only skills the user explicitly invokes via `/skill-name` run; auto-discovery via skill descriptions is disabled), `"name-only"` (skills are matched by exact name only; description-based auto-discovery is disabled). Use to keep a tighter rein on which skills the model loads or runs (v2.1.129) |
| `forceRemoteSettingsRefresh` | boolean | `false` | **(Managed only)** Block CLI startup until remote managed settings are freshly fetched. If the fetch fails, the CLI exits (fail-closed). Use in enterprise environments where policy enforcement must be up-to-date before any session begins (v2.1.92) |
| `wslInheritsWindowsSettings` | boolean | `false` | **(Windows managed settings only)** When `true`, Claude Code on WSL reads managed settings from the Windows policy chain (HKLM registry + `C:\Program Files\ClaudeCode\managed-settings.json`) in addition to `/etc/claude-code`, with Windows sources taking priority. Only honored when set in the HKLM registry key or `C:\Program Files\ClaudeCode\managed-settings.json`, both of which require Windows admin to write. For HKCU policy to also apply on WSL, the flag must additionally be set in HKCU itself. Has no effect on native Windows (v2.1.118) |
| `tui` | string | `"default"` | Rendering mode: `"fullscreen"` or `"default"`. Set via `/tui fullscreen` for flicker-free alt-screen rendering (v2.1.110) |
@@ -959,8 +960,13 @@ Set environment variables for all Claude Code sessions.
| `CLAUDE_CODE_FORCE_SYNC_OUTPUT` | Set to `1` to force-enable DEC private mode 2026 synchronized output when your terminal supports it but is not auto-detected. Useful for emulators such as Emacs `eat` that implement BSU/ESU but do not reply to the capability probe. Has no effect under tmux (v2.1.129) |
| `CLAUDE_CODE_SCROLL_SPEED` | Mouse wheel scroll multiplier for fullscreen rendering. Increase for faster scrolling, decrease for finer control |
| `CLAUDE_CODE_DISABLE_VIRTUAL_SCROLL` | Set to `1` to disable virtual scrolling in fullscreen rendering and render every message in the transcript. Use if scrolling in fullscreen mode shows blank regions where messages should appear |
| `CLAUDE_CODE_DISABLE_ALTERNATE_SCREEN` | Set to `1` to opt out of the alternate-screen (fullscreen) renderer entirely and use the classic scrollback renderer. Useful when terminal multiplexers, recording tools, or accessibility tooling do not handle the alt-screen buffer cleanly (v2.1.132) |
| `CLAUDE_CODE_DISABLE_MOUSE` | Set to `1` to disable mouse tracking in fullscreen rendering. Useful when mouse events interfere with terminal multiplexers or accessibility tools |
| `CLAUDE_CODE_HIDE_CWD` | Set to `1` to hide the current working directory in the Claude Code startup logo banner. Useful in screen recordings, demos, or shared sessions where the CWD path leaks information about the host or project layout (v2.1.119) |
| `CLAUDE_CODE_FORCE_SYNC_OUTPUT` | Set to `1` to force synchronous output flushing for Claude Code's writes to the terminal. Defaults to async/buffered output for performance. Use as a debugging aid when terminal output appears interleaved or out-of-order with subprocess output (v2.1.129) |
| `CLAUDE_CODE_PACKAGE_MANAGER_AUTO_UPDATE` | Control background package-manager-based auto-update checking for Claude Code. Set to `0` to disable the background check (Claude Code will not poll the package manager for newer versions); set to `1` (default) to keep the background check enabled. Independent of `DISABLE_AUTOUPDATER`, which gates the npm-registry auto-updater (v2.1.129) |
| `CLAUDE_CODE_ENABLE_GATEWAY_MODEL_DISCOVERY` | Set to `1` to opt into fetching the available-models list from the configured LLM gateway (e.g., a corporate proxy in front of Bedrock/Vertex). When enabled, the `/model` picker is populated from the gateway's discovery endpoint instead of the built-in alias list. Use when your gateway exposes a curated subset of models the user should choose from (v2.1.129) |
| `CLAUDE_CODE_ENABLE_FEEDBACK_SURVEY_FOR_OTEL` | Set to `1` to re-enable the in-session quality survey for OpenTelemetry-enabled enterprises. The survey is suppressed by default when `OTEL_*` env vars or `feedbackSurveyRate` are configured to avoid leaking survey traffic into enterprise telemetry pipelines. Use when admins want sampled survey data despite an OTel deployment (v2.1.136) |
| `CLAUDE_CODE_ACCESSIBILITY` | Set to `1` to keep native terminal cursor visible for screen readers and accessibility tools |
| `CLAUDE_CODE_SYNTAX_HIGHLIGHT` | Set to `0` to disable syntax highlighting in diff output |
| `CLAUDE_CODE_IDE_SKIP_AUTO_INSTALL` | Skip automatic IDE extension installation (`1` to skip) |
changelog/agent-collections/changelog.md
+46
View File
@@ -10,6 +10,52 @@ Tracks updates to the AGENT COLLECTIONS table in `README.md`.
---
## [2026-05-12 08:47 PM PKT] Agent Collections Update
| # | Priority | Type | Action | Status |
|---|----------|-------|-----------------------------------------------------------------------|-------------------------------------------------------------------------------------------------------------------------------------------------------|
| 1 | HIGH | Count | Update msitarzewski/agency-agents agents from 185 to 198 | COMPLETE (recursive tree scan: 198 agent .md files across 19 category dirs; +13 from 185; recent additions in specialized/ dir; NEW) |
| 2 | HIGH | Count | Update VoltAgent/awesome-claude-code-subagents agents from 145 to 189 | COMPLETE (recursive tree scan: 189 .md files under categories/, 10 READMEs excluded from 199 raw; +44 from 145; sustained PR activity ~5-8 agents/wk; NEW) |
| 3 | LOW | Sort | Verify sort order (stars descending) | COMPLETE (msitarzewski 96k > VoltAgent 20k — order preserved; RECURRING) |
---
## [2026-05-10 08:47 PM PKT] Agent Collections Update
| # | Priority | Type | Action | Status |
|---|----------|-------|----------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------------------------------------------------------|
| 1 | MED | Star | Update msitarzewski/agency-agents ★ from 95k to 96k | COMPLETE (HTML scrape ~95,700; crosses k-boundary; prev 95,300 → ~95,700; ~400 star increase) |
| 2 | MED | Star | Update VoltAgent/awesome-claude-code-subagents ★ from 19k to 20k | COMPLETE (HTML scrape ~19,500; crosses k-boundary; prev 19,433 → ~19,500; borderline half-k rounding applied) |
| 3 | LOW | Count | msitarzewski/agency-agents agents 185 → 175 (methodology diff) | INVALID (agent excluded strategy/playbooks(7)/runbooks(4)/coordination(2)/examples(5); different methodology from prior runs; RECURRING methodological variation; no change applied) |
| 4 | LOW | Count | VoltAgent/awesome-claude-code-subagents agents 145 → 144 | INVALID (RECURRING ±1 oscillation — 7th consecutive flip; per-category sum 144 unique (wordpress-master duplicate excluded); policy: no change at ±1) |
| 5 | LOW | Sort | Verify sort order (stars descending) | COMPLETE (msitarzewski 96k > VoltAgent 20k — order preserved) |
---
## [2026-05-09 08:46 PM PKT] Agent Collections Update
| # | Priority | Type | Action | Status |
|---|----------|-------|---------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------------------------------------------------------------------------|
| 1 | LOW | Count | Update VoltAgent/awesome-claude-code-subagents agents from 144 to 145 | COMPLETE (per-category enumeration across all 10 dirs: 145 .md files, conf 0.87; RECURRING 144↔145 oscillation — this is the 6th consecutive flip) |
| 2 | LOW | Star | msitarzewski/agency-agents ★ unchanged (95k = ~95,300) | INVALID (no change required) |
| 3 | LOW | Star | VoltAgent/awesome-claude-code-subagents ★ unchanged (19k = ~19,400) | INVALID (no change required) |
| 4 | LOW | Count | msitarzewski/agency-agents agents 185 vs reported 184 (−1) | INVALID (within ±1 margin of error; RECURRING oscillation; project-management/ truncation hint noted — 1-2 files may be missing from 184 count; no change) |
| 5 | LOW | Sort | Verify sort order (stars descending) | COMPLETE (msitarzewski 95k > VoltAgent 19k — order preserved) |
---
## [2026-05-09 06:57 PM PKT] Agent Collections Update
| # | Priority | Type | Action | Status |
|---|----------|-------|---------------------------------------------------------------------------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------|
| 1 | LOW | Count | Update msitarzewski/agency-agents agents from 185 to 186 | INVALID (within ±1 margin of error — Python crawl reported 186 but per-directory enumeration summed to 172; agent confidence 0.93; consistent with prior 184–186 oscillation policy) |
| 2 | LOW | Star | msitarzewski/agency-agents ★ unchanged (95k = 95,253) | INVALID (no change required) |
| 3 | LOW | Star | VoltAgent/awesome-claude-code-subagents ★ unchanged (19k = 19,433) | INVALID (no change required) |
| 4 | LOW | Count | VoltAgent/awesome-claude-code-subagents agents unchanged (144) | COMPLETE (tree not truncated; per-category sum exactly 144 across 10 dirs; prior 144↔145 oscillation definitively resolved as real content change, not pagination artifact) |
| 5 | LOW | Sort | Verify sort order (stars descending) | COMPLETE (msitarzewski 95k > VoltAgent 19k — order preserved) |
---
## [2026-05-08 08:46 PM PKT] Agent Collections Update
| # | Priority | Type | Action | Status |
changelog/best-practice/claude-settings/changelog.md
+33
View File
@@ -505,6 +505,39 @@
| 11 | LOW | Suspect Key Recurrence | `OTEL_LOG_TOOL_CONTENT` still changelog-only. Defer per Rule 8A | ✋ ON HOLD (kept — recurring from 2026-04-16 v2.1.110) |
| 12 | INVALID | Spurious Drift Claim | `workflow-claude-settings-agent` reported `autoSummaryEnabled` as a separate setting from `awaySummaryEnabled` (HIGH-confidence claim). Verified directly against official settings page — `autoSummaryEnabled` does NOT exist; only `awaySummaryEnabled` is documented | ❌ INVALID (agent contradicted by direct doc verification) |
| 13 | INVALID | Spurious Drift Claim | `workflow-claude-settings-agent` claimed `Agent` permission rule syntax should be `Agent(agent:name)` with an `agent:` prefix. Verified against official settings page — only references `Agent` rules without showing the `agent:` prefix syntax. Report's existing `Agent(name)` form matches established convention; no source confirms the prefix variant | ❌ INVALID (no source-verified evidence for `agent:` prefix) |
---
## [2026-05-09 06:58 PM PKT] Claude Code v2.1.138
| # | Priority | Type | Action | Status |
|---|----------|------|--------|--------|
| 1 | HIGH | Version Bump | Update report version badge from v2.1.126 → v2.1.138 and header "As of v2.1.126" → "As of v2.1.138" | ✅ COMPLETE (badge updated in Phase 2.6; body header line 6 updated to v2.1.138) |
| 2 | HIGH | New Setting | Add `worktree.baseRef` to Worktree Settings — string, values `"fresh"` or `"head"`, controls whether new worktrees branch from a fresh main HEAD or the current HEAD. Confirmed in v2.1.133 changelog | ✅ COMPLETE (added after `worktree.sparsePaths` with default `"fresh"` and v2.1.133 attribution) |
| 3 | HIGH | New Setting | Add `sandbox.bwrapPath` to Sandbox table — string (Linux/WSL managed-only), custom path to `bwrap` (bubblewrap) binary. Confirmed in v2.1.133 changelog | ✅ COMPLETE (added after `sandbox.enableWeakerNetworkIsolation` with managed-only annotation) |
| 4 | HIGH | New Setting | Add `sandbox.socatPath` to Sandbox table — string (Linux/WSL managed-only), custom path to `socat` binary. Confirmed in v2.1.133 changelog | ✅ COMPLETE (added after `sandbox.bwrapPath` with managed-only annotation) |
| 5 | HIGH | Changed Behavior | Update `autoMode` description to document the new `hard_deny` array — auto-mode classifier rules that block unconditionally, sibling to `allow` and `soft_deny`. Confirmed in v2.1.136 changelog | ✅ COMPLETE (description extended with `hard_deny` semantics, sentinel-incompatibility note, and v2.1.136 attribution) |
| 6 | HIGH | New Setting | Add `parentSettingsBehavior` (managed-only) to Settings Hierarchy section — string `"first-wins"` or `"merge"`, controls how SDK `managedSettings` parent tier merges. Confirmed in v2.1.133 changelog | ✅ COMPLETE (added to new "Dynamic & Parent-Tier Policy" subsection under Settings Hierarchy) |
| 7 | HIGH | New Setting | Add `policyHelper` (managed-only) to a managed-policy subsection — object with `path`, `timeoutMs`, `refreshIntervalMs`. Managed executable that computes managed settings dynamically. Confirmed in v2.1.136 changelog | ✅ COMPLETE (added to new "Dynamic & Parent-Tier Policy" subsection with field reference and use-case note) |
| 8 | HIGH | New Setting | Add `skillOverrides` to General Settings — string `"off"` / `"user-invocable-only"` / `"name-only"`, controls automatic skill invocation behavior. Confirmed in v2.1.129 changelog | ✅ COMPLETE (added after `disableSkillShellExecution` with all 3 enum values and v2.1.129 attribution) |
| 9 | HIGH | New Env Var | Add `CLAUDE_CODE_ENABLE_FEEDBACK_SURVEY_FOR_OTEL` to env vars table — re-enables session-quality feedback survey for OpenTelemetry-enabled enterprises. v2.1.136 changelog | ✅ COMPLETE (added in display/UI env vars cluster after `CLAUDE_CODE_ENABLE_GATEWAY_MODEL_DISCOVERY`) |
| 10 | HIGH | New Env Var | Add `CLAUDE_CODE_SESSION_ID` to env vars table — current session ID injected into Bash subprocess environment. v2.1.132 changelog | ✅ COMPLETE (added after `CLAUDECODE` with read-only annotation and v2.1.132 attribution) |
| 11 | HIGH | New Env Var | Add `CLAUDE_CODE_DISABLE_ALTERNATE_SCREEN` to env vars table — opt out of fullscreen renderer; use classic scrollback. v2.1.132 changelog | ✅ COMPLETE (added after `CLAUDE_CODE_DISABLE_VIRTUAL_SCROLL`) |
| 12 | HIGH | New Env Var | Add `CLAUDE_CODE_FORCE_SYNC_OUTPUT` to env vars table — force synchronous output (debugging aid). v2.1.129 changelog | ✅ COMPLETE (added after `CLAUDE_CODE_HIDE_CWD`) |
| 13 | HIGH | New Env Var | Add `CLAUDE_CODE_PACKAGE_MANAGER_AUTO_UPDATE` to env vars table — controls background package-manager-auto-update behavior. v2.1.129 changelog | ✅ COMPLETE (added after `CLAUDE_CODE_FORCE_SYNC_OUTPUT` with cross-reference to `DISABLE_AUTOUPDATER`) |
| 14 | HIGH | New Env Var | Add `CLAUDE_CODE_ENABLE_GATEWAY_MODEL_DISCOVERY` to env vars table — opt-in to fetching available models from LLM gateway. v2.1.129 changelog | ✅ COMPLETE (added after `CLAUDE_CODE_PACKAGE_MANAGER_AUTO_UPDATE`) |
| 15 | MED | New Env Var (Hook Input) | Add `$CLAUDE_EFFORT` to env vars table — Bash subprocess and hook input env exposing the active effort level (companion to `CLAUDE_CODE_EFFORT_LEVEL`). Hooks also receive `effort.level` JSON field. v2.1.133 changelog | ✋ ON HOLD (deferred — user scoped this run to HIGH priority items only; will pick up next run) |
| 16 | MED | Plugin Marketplace | Add brief note that plugin marketplace `source: 'settings'` is now supported for inline plugin entries directly in settings.json (v2.1.137 changelog) | ON HOLD (awaiting user approval) |
| 17 | MED | MCP Reserved Name | Add `workspace` to a "Reserved server names" note in MCP Servers section — v2.1.128 reserved this name for the workspace MCP integration | ON HOLD (awaiting user approval) |
| 18 | MED | Missing Setting | Add `disableRemoteControl` to Permissions/Managed-only settings — official permissions docs explicitly note "Remote Control can additionally be disabled per device with the `disableRemoteControl` managed setting" | ON HOLD (awaiting user approval) |
| 19 | MED | Missing Setting | Add `claudeMdExcludes` to Core Configuration — array, skip CLAUDE.md files matching globs. Listed on official settings page | ON HOLD (awaiting user approval) |
| 20 | MED | Missing Setting | Add `autoMemoryEnabled` to Core Configuration — boolean (default `true`), enables auto memory. Currently only `autoMemoryDirectory` is listed | ON HOLD (awaiting user approval) |
| 21 | MED | Example Update | Update Quick Reference example to showcase v2.1.129–v2.1.138 features — `worktree.baseRef`, `autoMode.hard_deny`, `skillOverrides`, plus a new env var | ON HOLD (awaiting user approval) |
| 22 | LOW | Header Count | Update header claim from "60+ settings" → "80+ settings" to better reflect actual official count after additions | ON HOLD (awaiting user approval) |
| 23 | LOW | Suspect Key Recurrence | `OTEL_LOG_TOOL_DETAILS` still "in v2.1.85 changelog, not yet on official env-vars page" after 17+ consecutive runs. Per Rule 10B, deferred pending official docs update | ON HOLD (kept — recurring from 2026-04-14 v2.1.107) |
| 24 | LOW | Suspect Key Recurrence | `OTEL_LOG_TOOL_CONTENT` still changelog-only. Defer per Rule 8A | ON HOLD (kept — recurring from 2026-04-16 v2.1.110) |
| 25 | INVALID | Spurious Drift Claim | `claude-code-guide` agent listed `disableSkillShellExecution` as NEW in v2.1.137. Verified against current report (line 89) — already documented in General Settings table | ❌ INVALID (already in report; pre-existing key) |
| 26 | INVALID | Spurious Drift Claim | `workflow-claude-settings-agent` flagged `syntaxHighlightingDisabled` as missing settings.json key. Could not confirm it as a `settings.json` key on official docs (the env var `CLAUDE_CODE_SYNTAX_HIGHLIGHT` is documented, the settings-key form was not source-verified). Per Rule 8A | ❌ INVALID (no source-verified evidence the settings.json key exists) |
| 14 | INVALID | Spurious Drift Claim | `workflow-claude-settings-agent` flagged `model` default (`"default"`) and `language` default (`"english"`) in Core Configuration as cosmetically wrong because the official docs show `-`. The report's values are descriptive placeholders explaining behavior when unset; flipping to `-` would lose information without any user-facing benefit | ❌ INVALID (cosmetic re-verification with no user-facing benefit) |
---