diff --git a/claude-code/changelog.html b/claude-code/changelog.html index ade2738..8272fc1 100644 --- a/claude-code/changelog.html +++ b/claude-code/changelog.html @@ -163,9 +163,19 @@

Template v1.2 2026-09-07 — As of CC v2.1.263Content / CC Release Updates

- October 2026 v2.1.287 – v2.1.295 6 releases + October 2026 v2.1.287 – v2.1.296 7 releases
+
+

CC v2.1.296 2026-10-09

+
    +
  • ADDautoCompactWindow — now also settable in subagent frontmatter and --agents definitions, so a subagent can auto-compact earlier than the main conversation's window
    • +
  • ADDCLAUDE_CODE_WORKFLOW_SUBAGENT_MODEL env var — run every Workflow tool agent on one model while other subagents keep theirs
    • +
  • ADDCLAUDE_CODE_OVERLOADED_RETRY_MAX_DELAY_MS env var — sets a longer maximum delay for the backoff when retrying an overloaded (529) request
    • +
  • CHGCLAUDE_CODE_MAX_MCP_DESCRIPTION_LENGTH — the default cap on MCP tool descriptions and server instructions is now 4,096 characters, raised from 2,048
    • +
+
+

CC v2.1.295 2026-10-08

    @@ -1575,7 +1585,7 @@

    Content Cards

- Template v1.2 · Content synced 2026-10-05
+ Template v1.2 · Content synced 2026-10-09
Unofficial community reference — not affiliated with or endorsed by Anthropic. Condensed from the official docs; always confirm against code.claude.com/docs.
diff --git a/claude-code/changes.json b/claude-code/changes.json index 2104b29..dac07b2 100644 --- a/claude-code/changes.json +++ b/claude-code/changes.json @@ -1,7 +1,35 @@ { - "generated": "2026-10-08", - "latest": "v2.1.295", + "generated": "2026-10-09", + "latest": "v2.1.296", "releases": [ + { + "version": "v2.1.296", + "label": "CC v2.1.296", + "date": "2026-10-09", + "id": "rel-2-1-296", + "entries": [ + { + "tag": "ADD", + "item": "autoCompactWindow", + "html": "autoCompactWindow — now also settable in subagent frontmatter and --agents definitions, so a subagent can auto-compact earlier than the main conversation's window" + }, + { + "tag": "ADD", + "item": "CLAUDE_CODE_WORKFLOW_SUBAGENT_MODEL", + "html": "CLAUDE_CODE_WORKFLOW_SUBAGENT_MODEL env var — run every Workflow tool agent on one model while other subagents keep theirs" + }, + { + "tag": "ADD", + "item": "CLAUDE_CODE_OVERLOADED_RETRY_MAX_DELAY_MS", + "html": "CLAUDE_CODE_OVERLOADED_RETRY_MAX_DELAY_MS env var — sets a longer maximum delay for the backoff when retrying an overloaded (529) request" + }, + { + "tag": "CHG", + "item": "CLAUDE_CODE_MAX_MCP_DESCRIPTION_LENGTH", + "html": "CLAUDE_CODE_MAX_MCP_DESCRIPTION_LENGTH — the default cap on MCP tool descriptions and server instructions is now 4,096 characters, raised from 2,048" + } + ] + }, { "version": "v2.1.295", "label": "CC v2.1.295", diff --git a/claude-code/feed.xml b/claude-code/feed.xml index 0a8719e..55c409f 100644 --- a/claude-code/feed.xml +++ b/claude-code/feed.xml @@ -4,7 +4,14 @@ https://dommango.github.io/agentmats/claude-code/ - 2026-10-08T09:00:00Z + 2026-10-09T09:00:00Z + + CC v2.1.296 — 4 placemat changes + https://dommango.github.io/agentmats/claude-code/changelog.html#rel-2-1-296 + + 2026-10-09T09:00:00Z + <ul><li><b>ADD</b> <code>autoCompactWindow</code> — now also settable in subagent frontmatter and <code>--agents</code> definitions, so a subagent can auto-compact earlier than the main conversation's window</li><li><b>ADD</b> <code>CLAUDE_CODE_WORKFLOW_SUBAGENT_MODEL</code> env var — run every Workflow tool agent on one model while other subagents keep theirs</li><li><b>ADD</b> <code>CLAUDE_CODE_OVERLOADED_RETRY_MAX_DELAY_MS</code> env var — sets a longer maximum delay for the backoff when retrying an overloaded (529) request</li><li><b>CHG</b> <code>CLAUDE_CODE_MAX_MCP_DESCRIPTION_LENGTH</code> — the default cap on MCP tool descriptions and server instructions is now 4,096 characters, raised from 2,048</li></ul> + CC v2.1.295 — 5 placemat changes https://dommango.github.io/agentmats/claude-code/changelog.html#rel-2-1-295 @@ -138,11 +145,4 @@ 2026-09-14T09:00:00Z <ul><li><b>CHG</b> <code>/fast [on/off]</code> — now works in Claude Code Remote sessions (cloud and self-hosted runners) too, via the host's fast-mode setting or by typing <code>/fast</code> in the session</li><li><b>CHG</b> <code>/config</code> — mouse support in fullscreen mode: the wheel scrolls the list, a click on a value changes it, and the row under the pointer is highlighted</li><li><b>ADD</b> <code>claude self-hosted-runner --drain-marker-file &lt;path&gt;</code> CLI flag — when present at a SIGTERM drain, the runner reports its exit to the server as a host drain (telemetry only)</li><li><b>ADD</b> <code>claude self-hosted-runner --host-config-snapshot disk|memory</code> CLI flag — choose where host config is cached; avoids silently losing it once the config directory exceeds 64 MiB</li><li><b>ADD</b> <code>allowed_domains</code> — per-command field on Bash, PowerShell and Monitor in sandboxed auto mode; the classifier reviews the domains a command needs and opens sandbox access to only those</li><li><b>ADD</b> <code>omitClaudeMd</code> agent frontmatter field (also settable via <code>--agents</code> JSON) — let a custom or plugin subagent run without user, project and local CLAUDE.md files; managed policy files still load</li><li><b>ADD</b> <code>claude plugin install</code> / <code>update</code> — added <code>--accept-command &lt;sha256&gt;</code> to accept exactly the command a previous <code>--json</code> run displayed, instead of <code>-y</code></li><li><b>CHG</b> <code>modelPricing</code> — the discount multiplier can now run above 1, up to 10, for marked-up internal chargeback rates; the Claude apps gateway <code>pricing</code> block supports the same range</li></ul> - - CC v2.1.269 — 6 placemat changes - https://dommango.github.io/agentmats/claude-code/changelog.html#rel-2-1-269 - - 2026-09-11T09:00:00Z - <ul><li><b>ADD</b> <code>claude plugin eval</code> CLI command — run a plugin's eval suite against Claude Code; outputs scored, reproducible results as JSON + HTML report</li><li><b>ADD</b> <code>/output-style [name]</code> slash command — list and switch output styles; works over Remote Control and in cloud and headless sessions</li><li><b>ADD</b> <code>bashEditDiffEnabled</code> setting — show a diff of files a Bash command changed in the tool result when the Bash tool handles file edits</li><li><b>ADD</b> <code>OTEL_METRICS_INCLUDE_REPOSITORY</code> env var — tag OpenTelemetry metrics and events with <code>vcs.*</code> repository attributes; commit events also emit <code>vcs.ref.head.*</code> with <code>OTEL_LOG_TOOL_DETAILS</code></li><li><b>ADD</b> <code>CLAUDE_CODE_GATEWAY_MODEL_DISCOVERY_TIMEOUT_MS</code> env var — extend the LLM gateway <code>/v1/models</code> discovery timeout (default 3 s)</li><li><b>ADD</b> <code>CLAUDE_CODE_WORKFLOW_MAX_CONCURRENT_AGENTS</code> env var — raise the Workflow tool's per-run concurrent agent limit (1–256) for inference-bound fan-outs</li></ul> - diff --git a/claude-code/index.html b/claude-code/index.html index 0abaad4..f8910d5 100644 --- a/claude-code/index.html +++ b/claude-code/index.html @@ -48,7 +48,7 @@

Claude Code PLACEMAT

Unofficial · auto-updated reference
- As of release: v2.1.295 + As of release: v2.1.296 Changelog → RSS @@ -85,7 +85,7 @@

Claude Code PLACEMAT

@@ -501,7 +501,7 @@

Environment Variables

#CLAUDECODEDetects if running inside CC shell; also injected into stdio MCP server subprocess environments #CLAUDE_CODE_SUBPROCESS_ENV_SCRUBStrip credentials from subprocesses; enables PID namespace isolation on Linux (sandboxing) #CLAUDE_CODE_MCP_SERVER_NAME / CLAUDE_CODE_MCP_SERVER_URLInject server identity into MCP headerHelper scripts - #CLAUDE_CODE_MAX_MCP_DESCRIPTION_LENGTHChange the 2,048-character cap on MCP tool descriptions and server instructions + #CLAUDE_CODE_MAX_MCP_DESCRIPTION_LENGTHChange the cap on MCP tool descriptions and server instructions #CLAUDE_CODE_NO_FLICKER=1Opt into flicker-free alt-screen rendering with virtualized scrollback #MCP_CONNECTION_NONBLOCKING=trueSkip MCP connection wait in headless (-p) mode #CLAUDE_CODE_PLUGIN_KEEP_MARKETPLACE_ON_FAILUREKeep marketplace cache when git pull fails (offline envs) @@ -565,6 +565,7 @@

Environment Variables

#FORCE_HYPERLINK=0Opt out of clickable hyperlinks in footer PR badge links (hyperlinks are on by default even when terminal support can't be detected, e.g. over ssh/tmux) #CLAUDE_CODE_WORKFLOW_PREFIX_STAGGER_MS=0Set to 0 to disable automatic staggering of same-prefix sibling workflow agents; staggering is on by default so later agents reuse the cached prompt prefix #CLAUDE_CODE_WORKFLOW_MAX_CONCURRENT_AGENTS Raise the Workflow tool's per-run concurrent agent limit (1–256) for inference-bound fan-outs + #CLAUDE_CODE_WORKFLOW_SUBAGENT_MODELRun every Workflow tool agent on one model while other subagents keep theirs #CLAUDE_CODE_TOOL_MEMORY_LIMITOpt-in memory cgroup limit for Bash tool commands on Linux; prevents a runaway build from stalling the session #CLAUDE_CODE_WEBFETCH_CACHE_TTL_MSConfigure the WebFetch session URL cache TTL in milliseconds (default: 15 minutes) #CLAUDE_CODE_WEBFETCH_DEADLINE_MS Override the 300-second WebFetch response deadline; set to 0 to disable the deadline entirely @@ -583,6 +584,7 @@

Environment Variables

#CLAUDE_CODE_NONSTREAMING_TIMEOUT_RETRIESCap re-sends of a non-streaming fallback request that timed out #CLAUDE_CODE_DISABLE_STRUCTURED_OUTPUTSTurn off structured outputs for gateways or Mantle setups that reject them, fixing session titles, memory recall and prompt hooks #CLAUDE_CODE_OVERLOADED_RETRY_BASE_DELAY_MSLengthen the base backoff delay used when retrying an overloaded (529) request + #CLAUDE_CODE_OVERLOADED_RETRY_MAX_DELAY_MSSet a longer maximum delay for the backoff when retrying an overloaded (529) request
@@ -649,6 +651,7 @@

Skills & Agent Frontmatter

#SendMessageMessage another agent or session; auto-resumes stopped agents #hooks:Agent lifecycle hooks; fire when running as main-thread agent via --agent #mcpServersMCP servers for the agent; now loaded for main-thread agent sessions via --agent + #autoCompactWindowSet when this subagent auto-compacts, independent of the main conversation's window @@ -747,7 +750,7 @@

Memory, Hooks & MCP

No matches — clear search to see everything.