Add cron bounded planning, skill path fixes, and color-coded agent logs#107
Merged
SergioChan merged 5 commits intomainfrom Mar 12, 2026
Merged
Conversation
The skill loader compacts absolute paths to ~/... notation via compactSkillPath, but path.resolve does not expand ~, causing the agent's read tool to resolve ~/.openpocket/skills/... relative to the workspace directory instead of the home directory. This resulted in ENOENT when the agent tried to read local skill files. Added expandTilde helper in resolveWorkspacePathPolicy so ~ paths resolve correctly to os.homedir(). Made-with: Cursor
The model was stripping or misinterpreting the ~ prefix in skill paths, using .openpocket/skills/... instead of ~/.openpocket/skills/..., which resolved to workspace/.openpocket/... and caused ENOENT. Using absolute paths eliminates the ambiguity entirely. Made-with: Cursor
Agent step logs ([thought], [decision], [result], etc.) were all plain white, making it hard to scan terminal output for important information. - Add colorizeAgentLog() that applies per-section ANSI colors: cyan for thought, yellow for decision, green/dim/red for result variants, dim for metadata sections (start/input/end) - Truncate multi-line result output to 4 lines with an omission summary, preventing skill reads and large outputs from flooding the terminal - Update runtimeTone to classify agent/model lines for the gateway log path as well Made-with: Cursor
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Why
Agent step logs (
[thought],[decision],[result],[start],[end]) were all plain white in the terminal, making it very hard to visually scan for important information. Skill file reads also flooded the terminal with dozens of lines of content.Changes
src/utils/cli-theme.ts: AddcolorizeAgentLog()with per-section ANSI colors (cyan for thought, yellow for decision, green/dim/red for result variants, dim for metadata). UpdateruntimeToneto classify[step]/[model]lines.src/agent/agent-runtime.ts: WirelogStepSectionthroughcolorizeAgentLog. Truncate multi-line result output to 4 lines + omission summary.src/agent/model-client.ts: Wire model decision/thought logs throughcolorizeAgentLog.src/skills/skill-loader.ts: Remove tilde shorthand from skill path resolution.src/agent/tool-policy.ts: Use absolute paths for skill locations.skills/x-twitter-play/SKILL.md: New skill definition.Testing
tsc --noEmit— compiles cleanly with no errorsnpm run build— dist output verified to containcolorizeAgentLogChecklist