🧾 docs: Refresh v0.8.7-rc1 Changelog (#609)

* docs: refresh v0.8.7-rc1 changelog

* Update token_usage.mdx

* Update config_v1.3.13.mdx

Author: danny-avila

content/changelog/config_v1.3.13.mdx

@@ -115,7 +115,7 @@ version: '1.3.13'
 
 - Added `interface.contextCost`
   - Controls whether usage cost is shown with context and token usage details
-  - Defaults to `true`
+  - Defaults to `false`; set to `true` to show usage costs in the UI
 
 - Added `interface.currency`
   - Converts displayed usage costs from USD into another currency with `code` and positive `rate`
@@ -125,6 +125,7 @@ version: '1.3.13'
   - Defines per-model context windows and per-million-token rates for custom endpoints
   - Supports `prompt`, `completion`, `context`, and optional `cacheRead` and `cacheWrite` values for each model name
   - Used by real-time context tracking, visible usage/cost breakdowns, and multi-endpoint agent billing
+  - Token-config cache entries are scoped by user when endpoint models, keys, URLs, or headers can vary by request context
 
 - Updated context usage after summarization
   - Persisted context usage now records a summary baseline marker after compaction
@@ -134,10 +135,11 @@ version: '1.3.13'
 - Added `skillSync.github`
   - Mirrors Skills from GitHub repositories into LibreChat
   - Supports scheduled syncs with `enabled`, `intervalMinutes`, `runOnStartup`, and `sources`
-  - Each source supports `id`, `owner`, `repo`, `ref`, `paths`, `skillDiscoveryDepth`, `credentialKey`, `token`, and `tenantId`
+  - Each source supports `id`, `owner`, `repo`, `ref`, `paths`, `skillDiscoveryDepth`, `credentialKey`, `token`
   - Sources must provide exactly one credential reference: either `credentialKey` for a stored admin credential or `token` as an environment variable reference like `${GITHUB_SKILLS_TOKEN}`
   - `skillDiscoveryDepth` defaults to `2` and is capped at `10`
   - `intervalMinutes` defaults to `60` and must be at least `5`
+  - Sync status responses are only shown to authorized admins
 
 - Added `messageFilter.pii`
   - Rejects submitted text that matches credential-shaped starter patterns or operator-defined regex patterns before it reaches the model or persistence
@@ -152,6 +154,7 @@ version: '1.3.13'
 - Updated MCP runtime placeholders
   - `{{LIBRECHAT_BODY_*}}` placeholders create request-scoped MCP connections that are reused for the current run and cleaned up when the request ends
   - Request-scoped MCP servers are excluded from the persistent tool cache so request-specific headers and URLs are not reused outside the active run
+  - Cleanup is tied to resumable run completion so reconnects and resumed streams keep request-scoped MCP resources alive until the run finishes
   - `{{LIBRECHAT_USER_*}}`, `{{LIBRECHAT_OPENID_*}}`, and `{{LIBRECHAT_GRAPH_*}}` placeholders still require user-scoped connections, but HTTP transports refresh resolved headers before each tool call without forcing a reconnect by themselves
 
 - Updated outbound proxy environment handling

content/changelog/v0.8.7-rc1.mdx

@@ -11,10 +11,10 @@ version: '0.8.7-rc1'
 
 - **Access Control, Auth, and MCP**
   - Added granular shared-link permissions, OpenID role sync, configurable OpenID token reuse windows, and MCP On-Behalf-Of token exchange for delegated downstream access.
-  - Hardened MCP OAuth prompts, token flow categorization, OpenID audience handling, shared-link caching, auth continuation limiting, ACL principal filtering, runtime user/request placeholders, request-scoped MCP connections, request-scoped MCP tool caching, outbound proxy handling, and configurable MCP OAuth completion windows.
+  - Hardened MCP OAuth prompts, token flow categorization, OpenID audience handling, shared-link caching, auth continuation limiting, ACL principal filtering, runtime user/request placeholders, request-scoped MCP connections, request-scoped MCP tool caching, resumable-run MCP cleanup, outbound proxy handling, and configurable MCP OAuth completion windows.
 - **Agents, Skills, and Projects**
   - Added private chat projects, agent file authoring tools, deployment-provided Skills, GitHub Skill Sync, model-spec Skill/Subagent scoping, model-spec conversation starters, soft-default model specs, model-spec landing branding, and the agent marketplace in the model selector.
-  - Deployment Skills now take precedence over persisted duplicates, memory-agent input is bounded, and repeated Skill bodies are deduplicated across fresh primes and message history.
+  - Deployment Skills now take precedence over persisted duplicates, Skill Sync status is scoped correctly, memory-agent input is bounded, and repeated Skill bodies are deduplicated across fresh primes and message history.
 - **Safety and Admin Controls**
   - Added configurable message PII filtering for credential-shaped text across caller-supplied API message roles and hardened model-spec Subagent config so clients cannot supply private Subagent allowlists.
 - **Observability and Operations**
@@ -23,9 +23,9 @@ version: '0.8.7-rc1'
   - Conversation titles can now generate immediately, message navigation supports focus and drag-to-scrub, message timestamps appear on hover, shared-link snapshots are sanitized, viewed branches stay stable when sibling branches change, file-authoring previews stream earlier, stale service-worker assets auto-recover after deploys, and reopened chats reload correctly from non-chat routes.
 - **Provider and Reasoning Updates**
   - Added Claude Fable 5, GPT-5.5, `chat-latest`, and refreshed frontier OpenAI defaults while dropping deprecated OpenAI defaults.
-  - Custom endpoints preserve reasoning params, support native Anthropic-compatible providers, Bedrock cache tokens no longer inflate completion counts, Gemini mixed tool config is enabled, Gemini/Vertex MCP tool schemas are sanitized, Imagen uses its configured region, built-in provider endpoints can forward custom gateway headers, and provider document uploads are preserved.
+  - Custom endpoints preserve reasoning params, support native Anthropic-compatible providers, withhold custom Anthropic headers from user-controlled URLs, Bedrock cache tokens no longer inflate completion counts, Gemini mixed tool config is enabled, Gemini/Vertex MCP tool schemas are sanitized, Imagen uses its configured region, built-in provider endpoints can forward custom gateway headers, and provider document uploads are preserved.
 - **Streaming and Performance**
-  - Completed Markdown blocks are memoized during streaming, tool execution can report per-call results through `onResult`, context usage and cost update in real time with persisted branch/total breakdowns and correct post-summarization baselines, subagent child-run usage is billed into parent transactions, model-spec icons survive stream resume/abort, HEIC conversion, locale resources, React Query Devtools, and Brotli asset serving are lazy/toggled, message tree streaming is covered by E2E tests, and shared package builds now use `tsdown` with isolated declarations.
+  - Completed Markdown blocks are memoized during streaming, tool execution can report per-call results through `onResult`, context usage updates in real time with persisted branch/total breakdowns and correct post-summarization baselines, cost display is opt-in by default, token-config caches are scoped correctly, subagent child-run usage is billed into parent transactions, model-spec icons survive stream resume/abort, HEIC conversion, locale resources, React Query Devtools, and Brotli asset serving are lazy/toggled, message tree streaming is covered by E2E tests, and shared package builds now use `tsdown` with isolated declarations.
 
 ---
 
@@ -143,6 +143,10 @@ version: '0.8.7-rc1'
 - 🔢 fix: Prevent "approximately" tildes from rendering as markdown subscript by [@danny-avila](https://github.com/danny-avila) in [#13743](https://github.com/danny-avila/LibreChat/pull/13743)
 - 🪙 fix: Correct Context Usage Gauge After Summarization by [@danny-avila](https://github.com/danny-avila) in [#13744](https://github.com/danny-avila/LibreChat/pull/13744)
 - 📈 fix: Isolate RUM Telemetry Proxy Auth from App Auth by [@danny-avila](https://github.com/danny-avila) in [#13765](https://github.com/danny-avila/LibreChat/pull/13765)
+- 🫷 fix: Withhold Anthropic Custom Headers From User URLs by [@danny-avila](https://github.com/danny-avila) in [#13767](https://github.com/danny-avila/LibreChat/pull/13767)
+- 🏘️ fix: Scope Skill Sync Status by [@danny-avila](https://github.com/danny-avila) in [#13771](https://github.com/danny-avila/LibreChat/pull/13771)
+- 🗂️ fix: Scope Token Config Cache by [@danny-avila](https://github.com/danny-avila) in [#13770](https://github.com/danny-avila/LibreChat/pull/13770)
+- 🪢 fix: Tie MCP Cleanup To Resumable Runs by [@danny-avila](https://github.com/danny-avila) in [#13769](https://github.com/danny-avila/LibreChat/pull/13769)
 
 ### 🔧 Refactoring and Performance
 
@@ -164,6 +168,7 @@ version: '0.8.7-rc1'
 - ⚙️ refactor: Brotli asset serving behind a feature toggle by [@upman](https://github.com/upman) in [#13641](https://github.com/danny-avila/LibreChat/pull/13641)
 - 🤫 refactor: Silent MCP OAuth Refresh on Mid-Session 401 by [@danny-avila](https://github.com/danny-avila) in [#13369](https://github.com/danny-avila/LibreChat/pull/13369)
 - 📡 refactor: Gate Noisy Redis OTEL Instrumentation by [@danny-avila](https://github.com/danny-avila) in [#13764](https://github.com/danny-avila/LibreChat/pull/13764)
+- 🧾 refactor: Disable Context Cost By Default by [@danny-avila](https://github.com/danny-avila) in [#13768](https://github.com/danny-avila/LibreChat/pull/13768)
 
 ### 🧪 Tests
 
@@ -208,12 +213,15 @@ version: '0.8.7-rc1'
 - 🧹 ci: Relieve disk pressure on GitNexus deploy by [@danny-avila](https://github.com/danny-avila) in [#13666](https://github.com/danny-avila/LibreChat/pull/13666)
 - 📦 chore: npm audit fix by [@danny-avila](https://github.com/danny-avila) in [#13701](https://github.com/danny-avila/LibreChat/pull/13701)
 - 🪟 fix: Cross-Platform Absolute-Path Check in tsdown neverBundle Predicates by [@danny-avila](https://github.com/danny-avila) in [#13700](https://github.com/danny-avila/LibreChat/pull/13700)
+- ✨ v0.8.7-rc1 by [@danny-avila](https://github.com/danny-avila) in [#13592](https://github.com/danny-avila/LibreChat/pull/13592)
+- 📊 chore: Bump Helm chart version to 2.0.6 by [@danny-avila](https://github.com/danny-avila) in [`5986a1c6d2`](https://github.com/danny-avila/LibreChat/commit/5986a1c6d2)
 
 ### 🌍 Internationalization
 
 - 🌍 i18n: Update translation.json with latest translations by [@danny-avila](https://github.com/danny-avila) in [#13466](https://github.com/danny-avila/LibreChat/pull/13466)
 - 🌍 i18n: Update translation.json with latest translations by [@danny-avila](https://github.com/danny-avila) in [#13482](https://github.com/danny-avila/LibreChat/pull/13482)
 - 🌍 i18n: Update translation.json with latest translations by [@danny-avila](https://github.com/danny-avila) in [#13505](https://github.com/danny-avila/LibreChat/pull/13505)
+- 🌍 i18n: Update translation.json with latest translations by [@danny-avila](https://github.com/danny-avila) in [#13766](https://github.com/danny-avila/LibreChat/pull/13766)
 
 ## New Contributors
 

content/docs/configuration/librechat_yaml/object_structure/interface.mdx

@@ -384,7 +384,12 @@ interface:
   ]}
 />
 
-**Default:** `true`
+**Default:** `false`
+
+**Notes:**
+
+- Set `contextCost: true` to opt in to visible cost values. Token usage can remain visible through `contextUsage` while cost values stay hidden.
+- `currency` only affects displayed costs when cost display is enabled.
 
 **Example:**
 

content/docs/configuration/token_usage.mdx

@@ -8,7 +8,7 @@ description: This covers how to track and control token usage in LibreChat. You
 
 As of `v0.6.0`, LibreChat accurately tracks token usage for supported endpoints. All token transactions are stored in the "Transactions" collection in your database. Current releases also show real-time context usage and cost in the conversation UI when enabled.
 
-Currently, you can limit user token usage by enabling user balances. Instead of configuring token credit limits via environment variables, you now set these options in your `librechat.yaml` file under the **balance** section.
+Currently, you can limit user token usage by enabling user balances. Instead of configuring token credit limits via environment variables, you now set these options in your `librechat.yaml` file under the **balance** section. Cost values are hidden by default and must be enabled with `interface.contextCost`.
 
 ## Viewing Context Usage and Cost
 
@@ -34,8 +34,8 @@ interface:
 ```
 
 - `contextUsage` controls whether users see the context window and token usage gauge.
-- `contextCost` controls whether users see cost values in the usage details.
-- `currency` converts displayed USD costs using a static multiplier. Transactions are still recorded using LibreChat's token credit accounting.
+- `contextCost` controls whether users see cost values in the usage details. It defaults to `false`; set it to `true` to show costs.
+- `currency` converts displayed USD costs using a static multiplier when cost display is enabled. Transactions are still recorded using LibreChat's token credit accounting.
 
 ## Custom Endpoint Token Config
 
@@ -58,6 +58,8 @@ endpoints:
 
 `prompt`, `completion`, and `context` are required for each model entry. `cacheRead` and `cacheWrite` can be added for providers that report cached input usage. For Agents that use multiple endpoints, LibreChat uses the matching endpoint/model token config when recording usage and cost.
 
+Fetched token config is cached with user scope when endpoint models, keys, URLs, or headers can vary by request context, so isolated custom endpoint pricing and context windows stay separated.
+
 ## Transaction Configuration
 
 The transaction system controls whether token usage records are saved to the database. This can be configured separately from the balance system.