@guidobuilds/forge-ai 0.1.0 → 0.3.0
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.
- package/CHANGELOG.md +60 -0
- package/README.md +14 -8
- package/{agents → artifacts/forge}/forge.md +15 -5
- package/artifacts/forge-grill/forge-grill.md +93 -0
- package/{skills/forge-worker/SKILL.md → artifacts/forge-worker/forge-worker.md} +22 -2
- package/{skills/using-forge/SKILL.md → artifacts/using-forge/using-forge.md} +1 -0
- package/dist/src/adapters/claude-known.js +31 -0
- package/dist/src/adapters/claude.js +33 -21
- package/dist/src/adapters/codex.js +14 -14
- package/dist/src/adapters/opencode.js +15 -15
- package/dist/src/adapters/shared.js +6 -1
- package/dist/src/cli.js +112 -24
- package/dist/src/discovery.js +16 -18
- package/dist/src/frontmatter.js +8 -68
- package/dist/src/manifest.js +134 -0
- package/dist/src/model.js +3 -0
- package/dist/src/processor.js +88 -45
- package/dist/src/writer.js +13 -1
- package/package.json +12 -12
- package/agents/forge-worker.md +0 -60
- package/skills/forge-build/SKILL.md +0 -80
- package/skills/forge-design/SKILL.md +0 -104
- package/skills/forge-explore/SKILL.md +0 -65
- package/skills/forge-helper/SKILL.md +0 -46
- package/skills/forge-plan/SKILL.md +0 -81
|
@@ -1,81 +0,0 @@
|
|
|
1
|
-
---
|
|
2
|
-
name: forge-plan
|
|
3
|
-
description: Create an execution plan from explore and design artifacts.
|
|
4
|
-
---
|
|
5
|
-
|
|
6
|
-
# Forge Plan Skill
|
|
7
|
-
|
|
8
|
-
## Role
|
|
9
|
-
Create the execution plan from the approved design work.
|
|
10
|
-
|
|
11
|
-
## Inputs
|
|
12
|
-
|
|
13
|
-
- `.forge/<feature-slug>/explore.md`
|
|
14
|
-
- `.forge/<feature-slug>/design.md`
|
|
15
|
-
|
|
16
|
-
## Required output file
|
|
17
|
-
|
|
18
|
-
`.forge/<feature-slug>/plan.md`
|
|
19
|
-
|
|
20
|
-
## Plan format
|
|
21
|
-
|
|
22
|
-
The plan defines execution order using building tasks.
|
|
23
|
-
|
|
24
|
-
A building task is a unit of work that must be implementable and testable.
|
|
25
|
-
|
|
26
|
-
The plan must include:
|
|
27
|
-
- Scope for the current delivery
|
|
28
|
-
- File Map covering the likely files or modules to touch and why
|
|
29
|
-
- Building tasks to execute
|
|
30
|
-
- Execution order of those building tasks
|
|
31
|
-
- Expected result of each building task
|
|
32
|
-
- Files or components touched by each building task
|
|
33
|
-
- Verification for each building task
|
|
34
|
-
- References to relevant `TASK-*` items when they clarify scope, sequencing, or dependencies
|
|
35
|
-
|
|
36
|
-
The plan must not include:
|
|
37
|
-
- Design definitions that belong in `design.md`
|
|
38
|
-
- Rationale for why a solution was chosen over another
|
|
39
|
-
- Code blocks
|
|
40
|
-
- Unresolved questions inside the plan document
|
|
41
|
-
- Placeholder language such as `TBD`, `TODO`, `implement later`, `adjust as needed`, or catch-all steps that hide concrete work
|
|
42
|
-
|
|
43
|
-
## Planning rules
|
|
44
|
-
|
|
45
|
-
- Use `design.md` as the source of truth for behavior and technical shape.
|
|
46
|
-
- Do not plan from a design artifact that still has unresolved critical decisions.
|
|
47
|
-
- Sequence delivery work without redefining the design.
|
|
48
|
-
- Make each building task buildable without guesswork.
|
|
49
|
-
- Prefer finer-grained tasks than the current format, but do not break work into trivial micro-steps.
|
|
50
|
-
- Ask questions only when uncertainty materially changes execution strategy or task ordering.
|
|
51
|
-
- A completed plan does not authorize implementation by itself.
|
|
52
|
-
- Default to `NEXT_RECOMMENDED: plan` when the plan is ready but waiting for user approval to build.
|
|
53
|
-
- Return `NEXT_RECOMMENDED: build` only when the orchestrator prompt explicitly states that the user has already approved implementation.
|
|
54
|
-
- Keep tasks surgically scoped to the approved goal; optional cleanup belongs outside the plan unless explicitly requested.
|
|
55
|
-
- The file map must justify why each listed file or module is expected to be touched.
|
|
56
|
-
- Prefer existing patterns and minimum necessary changes over broader structural rewrites.
|
|
57
|
-
- Each task must include minimal verification tied to the requested goal, not just a generic test step.
|
|
58
|
-
- If execution would require guesswork about assumptions, unknowns, or missing dependencies, stop and block instead of padding the plan with placeholders.
|
|
59
|
-
|
|
60
|
-
## Contract (strict)
|
|
61
|
-
|
|
62
|
-
Return only:
|
|
63
|
-
|
|
64
|
-
```text
|
|
65
|
-
STATUS: success|partial|blocked
|
|
66
|
-
PHASE: PLAN
|
|
67
|
-
FEATURE_SLUG: <kebab-case>
|
|
68
|
-
ARTIFACTS:
|
|
69
|
-
- .forge/<feature-slug>/plan.md
|
|
70
|
-
SUMMARY:
|
|
71
|
-
- <brief point>
|
|
72
|
-
NEXT_RECOMMENDED: plan|build
|
|
73
|
-
RISKS:
|
|
74
|
-
- <risk or None>
|
|
75
|
-
QUESTIONS:
|
|
76
|
-
1) <question>
|
|
77
|
-
2) <question>
|
|
78
|
-
```
|
|
79
|
-
|
|
80
|
-
Use `STATUS: blocked` when planning cannot continue due to missing critical decisions.
|
|
81
|
-
Include `QUESTIONS` only when blocked.
|