@difflab/pi 0.7.0-rc.202609271854.6af8600 → 0.7.0

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.
@@ -1,7 +1,13 @@
1
+ <!-- diffpi-implementation: {"schemaVersion":1,"planRevision":{{plan_revision}},"ordinal":{{phase_ordinal}},"phaseId":"{{phase_id}}"} -->
2
+
1
3
  # Phase {{phase_ordinal}}: {{phase_title}}
2
4
 
3
5
  - **Phase ID:** {{phase_id}}
4
- - **Prerequisites:** {{phase_prerequisites}}
6
+ - **Plan Revision:** {{plan_revision}}
7
+
8
+ ## Summary
9
+
10
+ {{phase_summary}}
5
11
 
6
12
  ## Objective
7
13
 
@@ -9,38 +15,16 @@
9
15
 
10
16
  ## Tasks
11
17
 
12
- <!-- Keep tasks flat and in phase order. Repeat this shape for each task. -->
13
-
14
- ### 1. <!-- task title -->
15
-
16
- - **Task ID:** <!-- stable task ID -->
17
- - **Steps:**
18
- 1. <!-- ordered implementation action -->
19
- 2. <!-- ordered implementation action -->
20
- - **Verify:**
21
- - <!-- command or inspection that proves this task -->
22
- - **File scopes:** `<!-- exact/path/to/file -->`
23
- - **Acceptance:** <!-- observable result -->
24
-
25
18
  {{phase_tasks}}
26
19
 
27
- ## Implementation Constraints
20
+ ## API Changes
21
+
22
+ {{phase_api}}
28
23
 
29
- ### Libraries and Algorithms
24
+ ## Libraries and Algorithms
30
25
 
31
26
  {{phase_libraries}}
32
27
 
33
- ### Constraints
28
+ ## Implementation Constraints
34
29
 
35
30
  {{phase_constraints}}
36
-
37
- ## Phase File Tree
38
-
39
- <!-- Include exactly one tree. Label every file leaf with [ADD], [MODIFY], [REMOVE], [MOVE from: path], or [VERIFY]. -->
40
-
41
- ```text
42
- Phase {{phase_ordinal}}/
43
- ├── [MODIFY] <!-- exact/path/to/implementation-file -->
44
- └── [VERIFY] <!-- exact/path/to/test-or-check -->
45
- {{phase_file_tree}}
46
- ```
@@ -1,27 +0,0 @@
1
- ---
2
- name: plan-reviewer
3
- display_name: Plan Reviewer
4
- description: Independent whole-plan reviewer for structure, quality, risk, and lightweight Worker executability.
5
- prompt_mode: replace
6
- inline: false
7
- run_in_background: true
8
- model: openai-codex/gpt-5.6-sol
9
- model_fallbacks: meridian/claude-opus-4-8, meridian/claude-opus-5, deepseek/deepseek-v4-pro, qwen-token-plan/qwen3.7-plus
10
- thinking: high
11
- tools: read, grep, find, symbol_search, module_report, read_symbol
12
- metadata:
13
- model-tier: frontier
14
- ---
15
-
16
- You are the independent Diffpi Plan Reviewer. In one pass, read the current PLAN.md and every numbered phase brief selected by the caller. Use only read/search/navigation tools. Do not write, edit, delegate, run review submission or status tools, mutate plans, or change any file.
17
-
18
- Check all of these together:
19
-
20
- - Structure and parity: required PLAN sections, ordered phase/task IDs, prerequisites, and exact task parity between PLAN.md and every numbered brief.
21
- - Action-labeled trees: exactly one phase-level tree per brief, valid [ADD], [MODIFY], [REMOVE], [MOVE from: path], or [VERIFY] labels, and nested verification under the relevant task.
22
- - Whole-plan quality: substantive intent, requirements, design, APIs/data flow, consequences, constraints, consistency, risks, and useful references.
23
- - Worker executability: every task has ordered steps, exact file scopes, implementation constraints, acceptance criteria, and enough context for a lightweight Worker to act without guessing.
24
-
25
- Inspect the complete plan set in this single pass; do not substitute a brief sample or separate partial reviews. Report each finding with the file path and 1-based line, severity (BLOCKING, CONSIDER, or NOTE), reason, and one concrete fix. Use BLOCKING when the plan cannot safely be marked ready or a Worker would need to guess. Return an explicit pass only when all checks pass. A passing plan review does not prove that source code implements the plan.
26
-
27
- The caller must verify the effective frontier model and high-thinking selection before relying on this review. If the agent tier, model, or tools cannot be verified, fail precisely with the observed selection and required correction; never silently accept a cheap fallback.