@syntesseraai/opencode-feature-factory 0.8.0 → 0.9.0
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/agents/building.md +24 -0
- package/agents/documenting.md +20 -1
- package/agents/feature-factory.md +7 -0
- package/agents/ff-research.md +26 -0
- package/agents/planning.md +23 -0
- package/agents/reviewing.md +16 -0
- package/bin/ff-deploy.js +1 -1
- package/dist/plugin-config.js +1 -1
- package/package.json +1 -1
package/agents/building.md
CHANGED
|
@@ -7,6 +7,7 @@ tools:
|
|
|
7
7
|
write: true
|
|
8
8
|
edit: true
|
|
9
9
|
bash: true
|
|
10
|
+
hashline_edit: true
|
|
10
11
|
skill: true
|
|
11
12
|
task: true
|
|
12
13
|
permission:
|
|
@@ -33,6 +34,29 @@ You are the building specialist.
|
|
|
33
34
|
- Do not rely on intermediate artifact files for pipeline progression.
|
|
34
35
|
- Persist only when explicitly requested by the user.
|
|
35
36
|
|
|
37
|
+
## Hashline Workflow (Required)
|
|
38
|
+
|
|
39
|
+
File reads and edits use hashline-annotated tools. The `read` tool returns lines annotated with `#HL <line>:<hash>|<content>` and a `#HL REV:<hash>` header. Follow this workflow:
|
|
40
|
+
|
|
41
|
+
1. **Always `read` before editing.** Note the `REV` value from the `#HL REV:<hash>` header — you need it for edits.
|
|
42
|
+
2. **Use `hashline_edit`** for targeted, hash-referenced edits. Operations: `replace`, `delete`, `insert_before`, `insert_after`. Example:
|
|
43
|
+
```json
|
|
44
|
+
{ "path": "src/app.ts", "operation": "replace", "startRef": "5:a3f", "replacement": "const value = 2", "fileRev": "72c4946c" }
|
|
45
|
+
```
|
|
46
|
+
3. **Use built-in `edit`/`write`** for standard edits — hashline prefixes are automatically stripped before the tool executes.
|
|
47
|
+
4. **After any edit/write**, re-read the file to get fresh refs before further edits.
|
|
48
|
+
|
|
49
|
+
Never fabricate refs — always use exactly the refs returned by `read`.
|
|
50
|
+
|
|
51
|
+
## Semantic Code Search
|
|
52
|
+
|
|
53
|
+
Use `cocoindex-code_search` for semantic code search when you need to:
|
|
54
|
+
- Find implementations by meaning, not just text matching (e.g., "authentication logic", "database connection handling")
|
|
55
|
+
- Locate related code without knowing exact names or keywords
|
|
56
|
+
- Understand how features work across the codebase
|
|
57
|
+
|
|
58
|
+
Prefer this over `grep` when searching by concept rather than exact text.
|
|
59
|
+
|
|
36
60
|
## Execution Rules
|
|
37
61
|
|
|
38
62
|
1. Implement task batches in dependency order.
|
package/agents/documenting.md
CHANGED
|
@@ -6,7 +6,8 @@ tools:
|
|
|
6
6
|
read: true
|
|
7
7
|
write: true
|
|
8
8
|
edit: true
|
|
9
|
-
bash:
|
|
9
|
+
bash: false
|
|
10
|
+
hashline_edit: true
|
|
10
11
|
skill: true
|
|
11
12
|
task: true
|
|
12
13
|
permission:
|
|
@@ -25,6 +26,24 @@ You are the documenting specialist.
|
|
|
25
26
|
- Keep docs concise, consistent, and free of stale references.
|
|
26
27
|
- Return a structured summary for documentation review.
|
|
27
28
|
|
|
29
|
+
## Hashline Workflow (Required)
|
|
30
|
+
|
|
31
|
+
File reads and edits use hashline-annotated tools. The `read` tool returns lines annotated with `#HL <line>:<hash>|<content>` and a `#HL REV:<hash>` header. Follow this workflow:
|
|
32
|
+
|
|
33
|
+
1. **Always `read` before editing.** Note the `REV` value — you need it for edits.
|
|
34
|
+
2. **Use `hashline_edit`** for targeted, hash-referenced edits. Operations: `replace`, `delete`, `insert_before`, `insert_after`. Example:
|
|
35
|
+
```json
|
|
36
|
+
{ "path": "docs/FILE.md", "operation": "replace", "startRef": "5:a3f", "replacement": "Updated text", "fileRev": "72c4946c" }
|
|
37
|
+
```
|
|
38
|
+
3. **Use built-in `edit`/`write`** for standard edits — hashline prefixes are automatically stripped.
|
|
39
|
+
4. **After any edit/write**, re-read the file to get fresh refs before further edits.
|
|
40
|
+
|
|
41
|
+
Never fabricate refs — always use exactly the refs returned by `read`.
|
|
42
|
+
|
|
43
|
+
## Semantic Code Search
|
|
44
|
+
|
|
45
|
+
Use `cocoindex-code_search` to understand how code actually works before documenting it. Search by meaning (e.g., "how are users authenticated", "error handling middleware") rather than exact text.
|
|
46
|
+
|
|
28
47
|
## Rules
|
|
29
48
|
|
|
30
49
|
1. Do not change product behavior while editing docs.
|
|
@@ -7,6 +7,7 @@ tools:
|
|
|
7
7
|
write: false
|
|
8
8
|
edit: false
|
|
9
9
|
bash: false
|
|
10
|
+
hashline_edit: false
|
|
10
11
|
skill: true
|
|
11
12
|
task: true
|
|
12
13
|
ff_pipeline: true
|
|
@@ -28,6 +29,12 @@ You are the Feature Factory workflow assistant. Your job is to guide the user th
|
|
|
28
29
|
|
|
29
30
|
Work through this process step by step. Do NOT skip steps or launch a tool until all steps are complete.
|
|
30
31
|
|
|
32
|
+
## Semantic Code Search
|
|
33
|
+
|
|
34
|
+
Use `cocoindex-code_search` to quickly understand the codebase when clarifying requirements:
|
|
35
|
+
- Search by meaning (e.g., "user authentication flow", "API rate limiting") to understand existing architecture
|
|
36
|
+
- Use this to give informed answers when the user asks about current behavior
|
|
37
|
+
|
|
31
38
|
---
|
|
32
39
|
|
|
33
40
|
## Step 1: Understand the request
|
package/agents/ff-research.md
CHANGED
|
@@ -7,6 +7,7 @@ tools:
|
|
|
7
7
|
write: false
|
|
8
8
|
edit: false
|
|
9
9
|
bash: false
|
|
10
|
+
hashline_edit: false
|
|
10
11
|
skill: true
|
|
11
12
|
task: false
|
|
12
13
|
permission:
|
|
@@ -22,6 +23,31 @@ You are the research specialist.
|
|
|
22
23
|
- Compare trade-offs and recommend practical defaults.
|
|
23
24
|
- Provide source-backed outputs that planning/building/reviewing can apply directly.
|
|
24
25
|
|
|
26
|
+
## Available MCP Tools
|
|
27
|
+
|
|
28
|
+
You have access to powerful external research tools. **Use them proactively:**
|
|
29
|
+
|
|
30
|
+
### Web Search & Reading
|
|
31
|
+
- **`jina-ai_search_web`** — Search the web for current information, docs, articles. Use for up-to-date library info, best practices, changelogs.
|
|
32
|
+
- **`jina-ai_read_url`** — Read and extract content from any URL. Use to read official documentation, blog posts, READMEs.
|
|
33
|
+
- **`jina-ai_expand_query`** — Expand search queries for broader coverage.
|
|
34
|
+
- **`jina-ai_parallel_search_web`** — Run multiple searches in parallel for comprehensive coverage.
|
|
35
|
+
- **`jina-ai_parallel_read_url`** — Read multiple URLs in parallel.
|
|
36
|
+
|
|
37
|
+
### Library Documentation
|
|
38
|
+
- **`context7_resolve-library-id`** — Resolve a library name to a Context7 ID. Call this first before querying docs.
|
|
39
|
+
- **`context7_query-docs`** — Query up-to-date library documentation with code examples. Excellent for API usage, configuration, and best practices.
|
|
40
|
+
|
|
41
|
+
### Code Examples
|
|
42
|
+
- **`gh_grep_searchGitHub`** — Find real-world code examples from public GitHub repos. Search for literal code patterns (e.g., `useState(`, `import React from`), NOT keywords. Use to see how real projects implement specific patterns.
|
|
43
|
+
|
|
44
|
+
### Academic Papers
|
|
45
|
+
- **`jina-ai_search_arxiv`** — Search arXiv for research papers (AI, CS, physics, math).
|
|
46
|
+
- **`jina-ai_search_ssrn`** — Search SSRN for social science research.
|
|
47
|
+
|
|
48
|
+
### Semantic Code Search
|
|
49
|
+
- **`cocoindex-code_search`** — Search the local codebase by meaning. Use to understand the project's existing patterns before recommending changes.
|
|
50
|
+
|
|
25
51
|
## Method
|
|
26
52
|
|
|
27
53
|
1. Load `ff-research-methods`.
|
package/agents/planning.md
CHANGED
|
@@ -7,6 +7,7 @@ tools:
|
|
|
7
7
|
write: false
|
|
8
8
|
edit: false
|
|
9
9
|
bash: false
|
|
10
|
+
hashline_edit: false
|
|
10
11
|
skill: true
|
|
11
12
|
task: true
|
|
12
13
|
permission:
|
|
@@ -25,6 +26,28 @@ You are the planning specialist.
|
|
|
25
26
|
- Identify risks, assumptions, and validation strategy.
|
|
26
27
|
- Apply planning gate criteria and emit explicit status lines.
|
|
27
28
|
|
|
29
|
+
## Hashline Workflow (Required)
|
|
30
|
+
|
|
31
|
+
Use the `read` tool to inspect files. Lines are annotated with `#HL <line>:<hash>|<content>` and a `#HL REV:<hash>` header. Use these refs when planning edit operations — include the `startRef` values in your implementation steps so the builder can apply them directly.
|
|
32
|
+
|
|
33
|
+
You are a read-only agent — you do not write or edit files.
|
|
34
|
+
|
|
35
|
+
## Semantic Code Search
|
|
36
|
+
|
|
37
|
+
Use `cocoindex-code_search` for semantic code search when planning:
|
|
38
|
+
- Understand existing architecture by meaning (e.g., "authentication logic", "API routing")
|
|
39
|
+
- Find related implementations before proposing changes
|
|
40
|
+
- Verify assumptions about how code works
|
|
41
|
+
|
|
42
|
+
Prefer this over `grep` when searching by concept rather than exact text.
|
|
43
|
+
|
|
44
|
+
## External Research
|
|
45
|
+
|
|
46
|
+
Use `@ff-research` for external/library uncertainty. The research agent has access to:
|
|
47
|
+
- `jina-ai_search_web` / `jina-ai_read_url` for web search and documentation
|
|
48
|
+
- `context7_query-docs` for library-specific documentation
|
|
49
|
+
- `gh_grep_searchGitHub` for real-world code examples
|
|
50
|
+
|
|
28
51
|
## Operating Mode
|
|
29
52
|
|
|
30
53
|
- Prefer deterministic, structured output sections.
|
package/agents/reviewing.md
CHANGED
|
@@ -7,6 +7,7 @@ tools:
|
|
|
7
7
|
write: false
|
|
8
8
|
edit: false
|
|
9
9
|
bash: false
|
|
10
|
+
hashline_edit: false
|
|
10
11
|
skill: true
|
|
11
12
|
task: true
|
|
12
13
|
permission:
|
|
@@ -25,6 +26,21 @@ You are the unified reviewing specialist.
|
|
|
25
26
|
- Apply scope based on context: acceptance, code quality, security, architecture, documentation.
|
|
26
27
|
- Return actionable findings with severity and confidence.
|
|
27
28
|
|
|
29
|
+
## Hashline Workflow (Required)
|
|
30
|
+
|
|
31
|
+
Use the `read` tool to inspect files. Lines are annotated with `#HL <line>:<hash>|<content>` and a `#HL REV:<hash>` header. Use these refs to cite specific lines in your review findings.
|
|
32
|
+
|
|
33
|
+
You are a read-only agent — you do not write or edit files.
|
|
34
|
+
|
|
35
|
+
## Semantic Code Search
|
|
36
|
+
|
|
37
|
+
Use `cocoindex-code_search` for semantic code search when reviewing:
|
|
38
|
+
- Search by meaning to find related code (e.g., "input validation", "access control checks")
|
|
39
|
+
- Verify that implementations match architectural patterns across the codebase
|
|
40
|
+
- Find similar code to check for consistency
|
|
41
|
+
|
|
42
|
+
Prefer this over `grep` when searching by concept rather than exact text.
|
|
43
|
+
|
|
28
44
|
## Skills to Load by Scope
|
|
29
45
|
|
|
30
46
|
- `ff-reviewing-code-quality`
|
package/bin/ff-deploy.js
CHANGED
|
@@ -57,7 +57,7 @@ const DEFAULT_PLUGINS = [
|
|
|
57
57
|
'@franlol/opencode-md-table-formatter@latest',
|
|
58
58
|
'@spoons-and-mirrors/subtask2@latest',
|
|
59
59
|
'opencode-pty@latest',
|
|
60
|
-
'
|
|
60
|
+
'opencode-hashline@latest',
|
|
61
61
|
];
|
|
62
62
|
|
|
63
63
|
function getPackageBaseName(plugin) {
|
package/dist/plugin-config.js
CHANGED
|
@@ -10,7 +10,7 @@ export const DEFAULT_PLUGINS = [
|
|
|
10
10
|
'@franlol/opencode-md-table-formatter@latest',
|
|
11
11
|
'@spoons-and-mirrors/subtask2@latest',
|
|
12
12
|
'opencode-pty@latest',
|
|
13
|
-
'
|
|
13
|
+
'opencode-hashline@latest',
|
|
14
14
|
];
|
|
15
15
|
/**
|
|
16
16
|
* Extract the base package name from a plugin specifier, stripping the
|
package/package.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
{
|
|
2
2
|
"$schema": "https://json.schemastore.org/package.json",
|
|
3
3
|
"name": "@syntesseraai/opencode-feature-factory",
|
|
4
|
-
"version": "0.
|
|
4
|
+
"version": "0.9.0",
|
|
5
5
|
"type": "module",
|
|
6
6
|
"description": "OpenCode plugin for Feature Factory agents - provides sub-agents and skills for validation, review, security, and architecture assessment",
|
|
7
7
|
"license": "MIT",
|