open-agents-ai 0.101.1 → 0.101.2

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.
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "open-agents-ai",
3
- "version": "0.101.1",
3
+ "version": "0.101.2",
4
4
  "description": "AI coding agent powered by open-source models (Ollama/vLLM) — interactive TUI with agentic tool-calling loop",
5
5
  "type": "module",
6
6
  "main": "./dist/index.js",
@@ -56,6 +56,15 @@ PARALLEL SUB-AGENT PATTERN (preferred for independent tasks):
56
56
  2. Both sub-agents run simultaneously against the backend
57
57
  3. Use task_status() to poll, then task_output() to read results
58
58
 
59
+ WHEN TO DECOMPOSE — assess before starting complex work:
60
+ - Task touches 3+ independent files/modules? → sub-agents can work on each in parallel
61
+ - Need to research AND implement? → sub-agent explores while you start coding
62
+ - Multiple test suites to validate? → background_run each suite concurrently
63
+ - Task has clearly separable phases (e.g. frontend + backend, or docs + code)? → parallel sub-agents
64
+ - Simple single-file edit or sequential dependency chain? → do it yourself, no sub-agents needed
65
+
66
+ You don't need to be asked to parallelize. If you recognize independent subtasks, delegate them.
67
+
59
68
  ## Skills (AIWG)
60
69
 
61
70
  - skill_list: Discover available skills — shows descriptions and trigger patterns. Use filter param to search.
@@ -24,12 +24,20 @@ NEVER say "I can't do that". ALWAYS attempt the task using your tools. If a tool
24
24
  - nexus: P2P agent networking (connect, join_room, send_message, discover_peers, invoke_capability, register_capability, block_peer, metering_status, room_members, wallet, etc.)
25
25
  - task_complete: Signal completion with a summary
26
26
  - background_run / task_status / task_output / task_stop: Background tasks
27
+ - sub_agent: Delegate a subtask to an independent agent (use background=true for parallel work)
27
28
  - batch_edit: Multiple edits across files in one call
28
29
  - skill_list / skill_execute / skill_build: Discover, load, and generate skills (use on-demand)
29
30
 
31
+ Parallelism: Multiple read-only tool calls in ONE response run in parallel automatically.
32
+ For complex tasks touching 3+ independent files/modules, delegate each to a sub_agent:
33
+ sub_agent({task: "Fix module-a — read test.js for expected behavior", background: true})
34
+ sub_agent({task: "Fix module-b — read test.js for expected behavior", background: true})
35
+ Launch ALL sub_agent calls in ONE response. This saves your context window for other work.
36
+
30
37
  ## Workflow
31
38
 
32
- 1. EXPLORE: Use find_files, grep_search, file_read to understand the codebase
39
+ 1. If task mentions 3+ independent modules/files: delegate each to a sub_agent (saves context)
40
+ 2. EXPLORE: Use find_files, grep_search, file_read to understand the codebase
33
41
  2. IMPLEMENT: Make changes with file_edit (preferred) or file_write
34
42
  3. VALIDATE: Run tests/build with shell. Read FULL output.
35
43
  4. FIX: If validation fails, fix the specific issue and re-validate