@iinm/plain-agent 1.10.17 → 1.10.18

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (2) hide show
  1. package/package.json +1 -1
  2. package/src/prompt.mjs +7 -10
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@iinm/plain-agent",
3
- "version": "1.10.17",
3
+ "version": "1.10.18",
4
4
  "description": "A lightweight coding agent for the terminal.",
5
5
  "license": "MIT",
6
6
  "type": "module",
package/src/prompt.mjs CHANGED
@@ -47,24 +47,21 @@ export function createPrompt({
47
47
  return `
48
48
  # Communication Style
49
49
 
50
- - Respond in the user's language.
51
- - Call the user by name, not "user".
52
- - Use emojis sparingly to highlight key points.
50
+ Respond in the user's language.
53
51
 
54
52
  # Memory Files
55
53
 
56
- - Create/Update memory files when creating/updating a plan, completing milestones, encountering issues, or making decisions.
57
- - Ensure self-containment: The file must be standalone. A reader should fully understand the task context, logic and progress without any other references.
54
+ Memory files preserve task state so work can be resumed after a context reset.
55
+
56
+ - Create/Update memory files when creating/updating a plan, completing milestones, encountering issues, or making decisions. Skip memory files for tasks that can be completed in a few steps.
57
+ - Ensure self-containment: Write as if the reader has no prior knowledge of the conversation.
58
58
  - Write the memory content in the user's language.
59
59
 
60
60
  Memory files should include:
61
- - Project discovery status: Whether AGENTS.md has been checked
62
61
  - Task overview: What the task is, why it's being done, requirements and constraints
63
- - Context: Relevant documentation, source files, commands, and resources referenced
62
+ - References: AGENTS.md, skills, relevant documentation, source files, and commands
64
63
  - Progress tracking: Completed milestones with evidence, current status, and next steps
65
- - Decision records: Important decisions made, alternatives considered, and rationale
66
- - Findings and learnings: Key discoveries, challenges encountered, and solutions applied
67
- - Future considerations: Known limitations, potential improvements, and follow-up items
64
+ - Decision records: Key decisions, alternatives considered, and rationale
68
65
 
69
66
  # Tools
70
67