{ "interval": { "intervalStart": "2026-01-19T00:00:00.000Z", "intervalEnd": "2026-01-20T00:00:00.000Z", "intervalType": "day" }, "repository": "elizaos/eliza", "overview": "From 2026-01-19 to 2026-01-20, elizaos/eliza had 1 new PRs (1 merged), 1 new issues, and 8 active contributors.", "topIssues": [ { "id": "I_kwDOMT5cIs7h2AmG", "title": "Fix TOCTOU race condition in credit deduction for streaming endpoints", "author": "linear", "number": 6338, "repository": "elizaos/eliza", "body": "Implement \"deduct-before, reconcile-after\" pattern to fix TOCTOU (Time-Of-Check-To-Time-Of-Use) race condition in streaming API endpoints.

Clean unification (we already do a good TOCTOU in MCP!
\n\n```markdown\nBEFORE (Vulnerable - Race Condition)\n════════════════════════════════════════════════════════════════════════════\n\n Request 1 ──┐\n Request 2 ──┼──▶ CHECK balance ≥ $0.02? ──▶ ✅ YES ──┐\n Request 3 ──┤ (all pass) │\n Request N ──┘ │\n ▼\n ┌──────────────────────────────────────────┐\n │ STREAMING (10+ sec) │\n │ No lock held on balance │\n └──────────────────────────────────────────┘\n │\n ▼\n All requests ──────────────────────────────▶ DEDUCT $0.02 each\n │\n ▼\n 💥 Balance goes NEGATIVE\n or deductions FAIL silently\n\n\nAFTER (Fixed - Atomic Reservation)\n════════════════════════════════════════════════════════════════════════════\n\n Request 1 ──┐ ┌─────────────────────────────────────────┐\n Request 2 ──┼──▶ │ ATOMIC: SELECT FOR UPDATE + DEDUCT │\n Request 3 ──┤ │ (PostgreSQL row lock on org balance) │\n Request N ──┘ └─────────────────────────────────────────┘\n │\n ┌─────────────────────────────┼─────────────────────────────┐\n │ │ │\n ▼ ▼ ▼\n Request 1 Request 2-N \n ✅ RESERVED ❌ 402 REJECTED\n $0.03 (with buffer) \"Insufficient credits\"\n │ \n ▼ \n ┌──────────────────────────────────────────┐\n │ STREAMING (10+ sec) │\n │ Credits already deducted │\n └──────────────────────────────────────────┘\n │\n ▼\n ┌──────────────────────────────────────────┐\n │ RECONCILE: actual cost = $0.012 │\n │ REFUND: $0.03 - $0.012 = $0.018 │\n └──────────────────────────────────────────┘\n │\n ▼\n ✅ Balance NEVER negative\n```", "createdAt": "2026-01-07T14:38:08Z", "closedAt": "2026-01-19T17:03:14Z", "state": "CLOSED", "commentCount": 0 }, { "id": "I_kwDOMT5cIs7kUu8H", "title": "Dashboard bug", "author": "borisudovicic", "number": 6382, "repository": "elizaos/eliza", "body": "Clicking out of the box here should take user back to original window when clicking on create again\n\n[Screen Recording 2026-01-19 at 18.29.34.mov](https://uploads.linear.app/186bdefa-3633-464a-80cd-6e86fe765a5c/bd077521-8d8c-4ae6-9876-9cdb11b952c6/b7864e34-b9cc-4f12-a690-ddf32af43f23?signature=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJwYXRoIjoiLzE4NmJkZWZhLTM2MzMtNDY0YS04MGNkLTZlODZmZTc2NWE1Yy9iZDA3NzUyMS04ZDhjLTRhZTYtOTg3Ni05Y2RiMTFiOTUyYzYvYjc4NjRlMzQtYjljYy00ZjEyLWE2OTAtZGRmMzJhZjQzZjIzIiwiaWF0IjoxNzY4ODQzODk3LCJleHAiOjE4MDA0MTQ0NTd9.8lPSAkcGHDqMzBoM8DMUpfqwiqf84cxSN-x5LEAeXZI)", "createdAt": "2026-01-19T17:31:38Z", "closedAt": null, "state": "OPEN", "commentCount": 0 } ], "topPRs": [ { "id": "PR_kwDOMT5cIs6xVNKQ", "title": "feat: Framework for adjusting prompts to best fix model contexts", "author": "odilitime", "number": 6113, "body": "# Risks\r\n\r\nLow, new feature, uses some memory but I put configurable limit on it\r\n\r\n# Background\r\n\r\n## What does this PR do?\r\n\r\nIt add a dynamic prompt execution system\r\n\r\n## What kind of change is this?\r\n\r\nFeatures (non-breaking change which adds functionality)\r\n\r\n## Why are we doing this? Any context or related work?\r\n\r\nPeople may use models with lower context and get hallucination, this helps detect when the context is blown out.\r\n\r\n# Documentation changes needed?\r\n\r\nMy changes require a change to the project documentation.\n\n\n---\n\n> [!NOTE]\n> Introduces a schema-based dynamic prompt executor with validation/retries/metrics and replaces ad-hoc prompting in message flows, plus supporting types/utils and test updates.\n> \n> - **Core Runtime**:\n> - Add `runtime.dynamicPromptExecFromState(...)` for schema-driven prompt execution (XML/JSON parsing, UUID validation codes, retries, Handlebars state injection, token estimations).\n> - Track execution metrics via `modelSchemaMetrics` and `modelMetrics` with optional LRU eviction (`DYNAMIC_PROMPT_MAX_ENTRIES`) and retry policy (`VALIDATION_LEVEL`).\n> - Expose metrics getters/clearers; helper methods for cache/keys; minor logging adjustments.\n> - **Message Service**:\n> - Replace manual `useModel` + XML parsing with `dynamicPromptExecFromState` in:\n> - `shouldRespond` evaluation (small model).\n> - Single-shot handler (requires `thought`, `actions`).\n> - Multi-step decision loop and final summary.\n> - Improve error logging and minor flow tweaks.\n> - **Types/Utils**:\n> - Export `SchemaRow` and `State` from `types`; extend `IAgentRuntime` with `dynamicPromptExecFromState`.\n> - Export `upgradeDoubleToTriple` and `composeRandomUser` from `utils`.\n> - **Tests/Test Utils**:\n> - Update `message-service` tests to mock `dynamicPromptExecFromState` responses.\n> - Add mock implementation in test runtime; minor logger formatting changes.\n> \n> Written by [Cursor Bugbot](https://cursor.com/dashboard?tab=bugbot) for commit 5905b5f9ab86dd7b7d727d92060575f42a8fc707. This will update automatically on new commits. Configure [here](https://cursor.com/dashboard?tab=bugbot).\n", "repository": "elizaos/eliza", "createdAt": "2025-11-04T05:53:02Z", "mergedAt": null, "additions": 3464, "deletions": 329 }, { "id": "PR_kwDOMT5cIs6-GhMD", "title": "[Draft] RLM provider prototype for Eliza Python core", "author": "matomoniwano", "number": 6383, "body": "# Relates to\r\n\r\nN/A (exploratory / draft integration)\r\n\r\n# Risks\r\n\r\nLow\r\n\r\n- Provider is opt-in and not registered by default\r\n- No existing code paths are modified unless the RLM provider is explicitly used\r\n- Fallback stub behavior when RLM dependency is not installed\r\n\r\n# Background\r\n\r\n## What does this PR do?\r\n\r\n\r\nThis PR introduces a **prototype RLM provider** for the Eliza Python core.\r\n\r\nIt adds:\r\n- A minimal `RLMProvider` that follows the existing provider interface\r\n- A thin `RLMClient` adapter that wraps an external Recursive Language Model (RLM)\r\n- Safe stub behavior when the RLM dependency is not present\r\n\r\nThe goal is to explore whether RLM can be integrated cleanly as a reasoning backend,\r\nwithout affecting Eliza’s memory, planning, or autonomy layers.\r\n\r\nThis is intentionally an early draft meant for discussion and iteration.\r\n\r\n## What kind of change is this?\r\n\r\nFeatures (non-breaking, opt-in)\r\n\r\n## Why are we doing this? Any context or related work?\r\n\r\nIdea originated from this tweet https://x.com/shawmakesmagic/status/2013151880810057772?s=20\r\n\r\nThis PR is an experiment to see if RLM can fit naturally into the existing\r\nprovider/model boundary without increasing coupling or complexity.\r\n\r\n# Documentation changes needed?\r\n\r\nMy changes do not require a change to the project documentation.\r\n\r\n# Testing\r\n\r\n## Where should a reviewer start?\r\n\r\n- `elizaos/providers/rlm_provider.py`\r\n- `elizaos/providers/rlm_client.py`\r\n- `elizaos/providers/README.md`\r\n- `python/tests/test_rlm_provider.py`\r\n\r\n## Detailed testing steps\r\n\r\n- Instantiate `RLMProvider`\r\n- Call `generate_text` with a prompt or messages\r\n- Verify stub behavior when RLM is not installed\r\n- Verify successful pass-through when RLM is available locally\r\n\r\nNo automated tests are included at this stage.\r\n\r\n# Deploy Notes\r\n\r\nn/a\r\n\r\n## Discord username\r\nDiscord: matomo8925\r\nX: https://x.com/momo_mattomo\r\nPrj X: https://x.com/AgentRLM", "repository": "elizaos/eliza", "createdAt": "2026-01-19T23:52:38Z", "mergedAt": null, "additions": 387, "deletions": 0 }, { "id": "PR_kwDOMT5cIs69wY0l", "title": "feat: add unregisterAction method to core runtime", "author": "YuriNachos", "number": 6372, "body": "## Summary\n- Adds ability to unregister actions from the runtime\n- Implements `unregisterAction` method using array mutation (splice)\n- Adds comprehensive test coverage for unregister functionality\n\n## Problem\nThere was no way to remove actions from the runtime once they were registered. This made it impossible to dynamically manage actions during runtime.\n\n## Solution\nAdded `unregisterAction(actionName: string): boolean` method to:\n- `IAgentRuntime` interface in `packages/core/src/types/runtime.ts`\n- Runtime implementation in `packages/core/src/runtime.ts`\n- Test coverage in `packages/core/src/__tests__/runtime.test.ts`\n\nThe implementation uses `splice()` to modify the readonly array in-place, avoiding TypeScript errors while maintaining readonly semantics for external consumers.\n\n## Changes\n- `packages/core/src/types/runtime.ts`: Added `unregisterAction` method signature\n- `packages/core/src/runtime.ts`: Implemented `unregisterAction` using array mutation\n- `packages/core/src/__tests__/runtime.test.ts`: Added 3 test cases\n\nFixes #5697\n\n---\n\n🤖 Generated with [Claude Code](https://claude.com/claude-code)\n\n\n\n

Greptile Summary

\n\n\nThis PR adds the ability to dynamically unregister actions from the runtime, addressing issue #5697. The implementation includes a new `unregisterAction` method that removes actions by name and returns a boolean indicating success. The PR also includes an unrelated but beneficial fix to the reflection evaluator that ensures entity connections exist before saving facts, preventing potential database constraint violations.\n\n**Key Changes:**\n- Added `unregisterAction(actionName: string): boolean` to the `IAgentRuntime` interface\n- Implemented the method in `AgentRuntime` using `Array.splice()` with proper logging (debug on success, warn on not found)\n- Added 3 comprehensive test cases covering successful unregistration, non-existent action handling, and selective removal\n- Enhanced reflection evaluator to call `ensureConnection` before saving facts, with proper error handling\n- Added test coverage for the reflection evaluator entity connection logic\n- Changed null check from `!=` to `!==` for stricter equality (minor code quality improvement)\n\n**Implementation Quality:**\nThe `unregisterAction` implementation follows the project's patterns with appropriate logging and clean boolean return semantics. The use of `splice()` on the readonly array is acceptable as it maintains external immutability while allowing internal mutation. The test coverage is thorough and follows existing test patterns using `bun:test`.\n\n**Reflection Evaluator Enhancement:**\nThe plugin-bootstrap changes are marked as part of this PR but appear to be a separate logical fix (ensuring entity connections before fact persistence). This is a positive addition that prevents runtime errors, though it could have been a separate PR for clearer change tracking.\n\n

Confidence Score: 5/5

\n\n\n- This PR is safe to merge with minimal risk\n- The implementation is straightforward, well-tested, and follows existing patterns in the codebase. The `unregisterAction` method uses standard array manipulation with proper error handling and logging. All 3 test cases pass and cover the main scenarios. The reflection evaluator enhancement is a defensive improvement that prevents potential database errors. No breaking changes or security concerns identified.\n- No files require special attention\n\n

Important Files Changed

\n\n\n\n\n| Filename | Overview |\n|----------|----------|\n| packages/core/src/types/runtime.ts | Added `unregisterAction` method signature to `IAgentRuntime` interface, consistent with existing patterns |\n| packages/core/src/runtime.ts | Implemented `unregisterAction` method with proper logging and boolean return value indicating success/failure |\n| packages/core/src/__tests__/runtime.test.ts | Added comprehensive test coverage for unregister functionality with 3 test cases covering happy path, error case, and selective removal |\n| packages/plugin-bootstrap/src/evaluators/reflection.ts | Enhanced reflection evaluator to ensure entity connections before saving facts, preventing database constraint violations |\n| packages/plugin-bootstrap/src/__tests__/reflection-entity-connection.test.ts | Added test coverage for reflection evaluator entity connection logic |\n\n\n\n\n\n

Sequence Diagram

\n\n```mermaid\nsequenceDiagram\n participant Client\n participant Runtime\n participant Actions Array\n participant Logger\n\n Note over Client,Logger: unregisterAction Flow\n\n Client->>Runtime: unregisterAction(\"actionName\")\n Runtime->>Actions Array: findIndex(a => a.name === \"actionName\")\n \n alt Action Found (index !== -1)\n Actions Array-->>Runtime: return valid index\n Runtime->>Actions Array: splice(index, 1)\n Actions Array-->>Runtime: action removed\n Runtime->>Logger: debug(\"Action unregistered\")\n Runtime-->>Client: return true\n else Action Not Found (index === -1)\n Actions Array-->>Runtime: return -1\n Runtime->>Logger: warn(\"Action not found, cannot unregister\")\n Runtime-->>Client: return false\n end\n\n Note over Client,Logger: Reflection Evaluator Enhancement\n\n participant Evaluator\n participant State\n participant RoomDB\n\n Evaluator->>Runtime: Extract facts from AI response\n Evaluator->>State: Get room from state?.data?.room\n \n alt Room not in state\n Evaluator->>RoomDB: getRoom(roomId)\n RoomDB-->>Evaluator: return room with worldId\n end\n \n alt Room or worldId missing\n Evaluator->>Logger: warn(\"Room or world ID not found\")\n Evaluator-->>Evaluator: return early (skip fact save)\n else Room and worldId exist\n Evaluator->>Runtime: ensureConnection({entityId, roomId, worldId, ...})\n alt Connection Success\n Runtime-->>Evaluator: connection ensured\n Evaluator->>Logger: debug(\"Entity connection ensured\")\n Evaluator->>Runtime: createMemory(fact)\n Runtime-->>Evaluator: fact saved\n else Connection Failure\n Runtime-->>Evaluator: throw error\n Evaluator->>Logger: error(\"Failed to ensure entity connection\")\n Evaluator-->>Evaluator: return early (skip fact save)\n end\n end\n```\n\n\n\n", "repository": "elizaos/eliza", "createdAt": "2026-01-17T08:13:14Z", "mergedAt": null, "additions": 207, "deletions": 2 }, { "id": "PR_kwDOMT5cIs68O-Yj", "title": "chore: Optimize build task inputs in turbo.json", "author": "odilitime", "number": 6349, "body": "Add explicit inputs to build task for cache optimization\r\n\r\n# Risks\r\n\r\nLow\r\n\r\n# Background\r\n\r\n## What does this PR do?\r\n\r\nMake turbo rebuild less\r\n\r\n## What kind of change is this?\r\n\r\nImprovements (misc. changes to existing features)\r\n\r\n## Why are we doing this? Any context or related work?\r\n\r\nMy turbo was eating over 40gb of ram on some changes\r\n\r\n# Documentation changes needed?\r\n\r\nMy changes do not require a change to the project documentation.\n\n\n---\n\n> [!NOTE]\n> Optimizes Turbo caching by narrowing build task inputs and adding global dependencies, reducing unnecessary rebuilds without affecting runtime code.\n> \n> - Adds explicit `inputs` for `build`, `@elizaos/client#build`, `@elizaos/cli#build`, and `@elizaos/server#build` to hash only build-relevant files (e.g., `src/**`, `tsconfig*.json`, `vite.config.ts`, `bunfig.toml`)\n> - Introduces `globalDependencies` (`tsconfig.json`, `tsconfig.build.template.json`) to ensure consistent cache invalidation across the repo\n> - Outputs remain `dist/**`; other task behavior unchanged\n> \n> Written by [Cursor Bugbot](https://cursor.com/dashboard?tab=bugbot) for commit df4ea5700066607c79dccc2f0c7125d4813a2cf0. This will update automatically on new commits. Configure [here](https://cursor.com/dashboard?tab=bugbot).\n\n\n\n\n

Greptile Overview

\n\n

Greptile Overview

\n\n### Greptile Summary\n\nOptimizes Turbo build cache by adding explicit `inputs` to narrow cache invalidation to build-relevant files. This prevents unnecessary rebuilds when documentation, tests, or other non-source files change, addressing the author's RAM consumption issue. The change includes helpful inline comments explaining the optimization rationale.\n\n### Confidence Score: 3/5\n\n- Safe to merge with suggested additions to prevent cache misses\n- The optimization is sound and addresses a real performance issue. However, the inputs list is incomplete—missing `bunfig.toml` and `eslint.config.js` could cause cache hits when these files change, leading to stale builds. These files exist across multiple packages and can affect build behavior. With the suggested additions, this would be a solid optimization.\n- turbo.json needs additional build-relevant config files in the inputs list\n\n

Important Files Changed

\n\n\n\nFile Analysis\n\n\n\n| Filename | Score | Overview |\n|----------|-------|----------|\n| turbo.json | 3/5 | Adds explicit inputs to build task to optimize cache invalidation; missing bunfig.toml and eslint.config.js which can affect builds |\n\n\n\n\n\n

Sequence Diagram

\n\n```mermaid\nsequenceDiagram\n participant Dev as Developer\n participant Turbo as Turbo Cache\n participant Build as Build Process\n \n Dev->>Turbo: \"Change README.md\"\n Note over Turbo: Before: Hash ALL files
Cache MISS → Rebuild\n Dev->>Turbo: \"Change README.md (with inputs)\"\n Note over Turbo: After: Hash only inputs
Cache HIT → Skip rebuild\n \n Dev->>Turbo: \"Change src/index.ts\"\n Turbo->>Build: \"Cache MISS\"\n Build-->>Turbo: \"Build complete\"\n Turbo-->>Dev: \"New cache entry\"\n```\n\n

Important Files Changed

\n\n\n\nFile Analysis\n\n\n\n| Filename | Score | Overview |\n|----------|-------|----------|\n| turbo.json | 3/5 | Adds explicit inputs to build task to optimize cache invalidation; missing bunfig.toml and eslint.config.js which can affect builds |\n\n\n\n\n\n

Sequence Diagram

\n\n```mermaid\nsequenceDiagram\n participant Dev as Developer\n participant Turbo as Turbo Cache\n participant Build as Build Process\n \n Dev->>Turbo: \"Change README.md\"\n Note over Turbo: Before: Hash ALL files
Cache MISS → Rebuild\n Dev->>Turbo: \"Change README.md (with inputs)\"\n Note over Turbo: After: Hash only inputs
Cache HIT → Skip rebuild\n \n Dev->>Turbo: \"Change src/index.ts\"\n Turbo->>Build: \"Cache MISS\"\n Build-->>Turbo: \"Build complete\"\n Turbo-->>Dev: \"New cache entry\"\n```\n\n\n\n\n\n\n\n## Summary by CodeRabbit\n\n* **Chores**\n * Optimized build caching configuration to improve build performance and reduce unnecessary cache invalidation across the monorepo.\n\n✏️ Tip: You can customize this high-level summary in your review settings.\n\n", "repository": "elizaos/eliza", "createdAt": "2026-01-09T00:52:34Z", "mergedAt": "2026-01-19T19:48:17Z", "additions": 47, "deletions": 0 } ], "codeChanges": { "additions": 47, "deletions": 0, "files": 1, "commitCount": 12 }, "completedItems": [ { "title": "chore: Optimize build task inputs in turbo.json", "prNumber": 6349, "type": "other", "body": "Add explicit inputs to build task for cache optimization\r\n\r\n# Risks\r\n\r\nLow\r\n\r\n# Background\r\n\r\n## What does this PR do?\r\n\r\nMake turbo rebuild less\r\n\r\n## What kind of change is this?\r\n\r\nImprovements (misc. changes to existing features)\r\n\r\n## ", "files": [ "turbo.json" ] } ], "topContributors": [ { "username": "matomoniwano", "avatarUrl": "https://avatars.githubusercontent.com/u/47988393?u=2e31304db3ca7b0a1f62bc26443c25ec34bb519d&v=4", "totalScore": 29.69251334905818, "prScore": 29.69251334905818, "issueScore": 0, "reviewScore": 0, "commentScore": 0, "summary": "matomoniwano: Focused on developing a new RLM provider prototype for Eliza Python core, evidenced by the open PR elizaos/eliza#6383, which involved modifying 16 files with significant changes (+610/-223 lines) primarily across configuration, documentation, and tests." }, { "username": "greptile-apps", "avatarUrl": "https://avatars.githubusercontent.com/in/867647?v=4", "totalScore": 13.5, "prScore": 0, "issueScore": 0, "reviewScore": 13.5, "commentScore": 0, "summary": "greptile-apps: No activity today." }, { "username": "shuhaib112", "avatarUrl": "https://avatars.githubusercontent.com/u/211030292?v=4", "totalScore": 9, "prScore": 0, "issueScore": 0, "reviewScore": 9, "commentScore": 0, "summary": "shuhaib112: No activity today." }, { "username": "ChristopherTrimboli", "avatarUrl": "https://avatars.githubusercontent.com/u/27584221?u=0d816ce1dcdea8f925aba18bb710153d4a87a719&v=4", "totalScore": 5, "prScore": 0, "issueScore": 0, "reviewScore": 5, "commentScore": 0, "summary": "ChristopherTrimboli: Today's activity included modifying 30 files with a near-even split of additions and deletions (+219/-216 lines) across 6 commits, primarily focused on other work (67%) and bug fixes (33%), and provided one approval review." }, { "username": "borisudovicic", "avatarUrl": "https://avatars.githubusercontent.com/u/31806472?u=8935f4d43fd7e4eb9bf5ff92d54d4d2f8ac8a786&v=4", "totalScore": 2, "prScore": 0, "issueScore": 2, "reviewScore": 0, "commentScore": 0, "summary": "borisudovicic: Focused on identifying and reporting potential issues, creating one issue in elizaos/eliza (#6382) to report a \"Dashboard bug.\"" }, { "username": "odilitime", "avatarUrl": "https://avatars.githubusercontent.com/u/16395496?u=c9bac48e632aae594a0d85aaf9e9c9c69b674d8b&v=4", "totalScore": 0.33999999999999997, "prScore": 0, "issueScore": 0, "reviewScore": 0, "commentScore": 0.33999999999999997, "summary": null } ], "newPRs": 1, "mergedPRs": 1, "newIssues": 1, "closedIssues": 1, "activeContributors": 8 }