1# Research Thread: [Run ID]
2 
3## Mission
4 
5- Objective:
6- Scope:
7- Started:
8- Owner:
9- Current status: active / paused / complete / abandoned
10 
11## Locked Surfaces
12 
13List files or rules the agent may read but not mutate during this run.
14 
15- `researcher/rubrics/content-curation.md`
16- `researcher/rubrics/skill-change.md`
17- `researcher/rubrics/harness-change.md`
18- `researcher/mechanisms/registry.jsonl`
19- `.claude-plugin/marketplace.json`
20- `.plugin/plugin.json`
21- Merge policy: agents may prepare PRs, but human approval is required for push and merge.
22 
23## Editable Surfaces
24 
25List files the agent may create or update during this run.
26 
27- Source evaluations:
28- Skill proposals:
29- Mechanism proposals:
30- Run state:
31- Draft branches or patches:
32 
33## Source Queue
34 
35| ID | Source | Status | Next Action |
36| --- | --- | --- | --- |
37| S001 |  | discovered | evaluate |
38 
39## Decisions
40 
41Append decisions chronologically.
42 
43```text
44T+00:00 decision: [what changed]
45reason: [why]
46evidence: [file or URL]
47next: [next action]
48```
49 
50## Experiments And Evaluations
51 
52| ID | Artifact | Rubric | Result | Notes |
53| --- | --- | --- | --- | --- |
54| E001 |  |  |  |  |
55 
56## Open Questions
57 
58- 
59 
60## Handover Summary
61 
62Use this section whenever context compacts or another agent resumes the run.
63 
64- Best current candidate:
65- Rejected candidates:
66- Unresolved risks:
67- Files to read first:
68- Next concrete action:
69