brainclaw 0.19.7 → 0.19.10

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.
@@ -1,99 +1,131 @@
1
1
  # Quickstart
2
2
 
3
- This guide walks through the shortest path to getting value from brainclaw.
4
-
5
- ## Important limitation for now
6
-
7
- Do not run multiple coding agents in parallel on the same project checkout yet.
8
-
9
- brainclaw is already useful for sequential collaboration: one agent can pick up where another stopped, inspect shared context, and continue from explicit plans, claims, traps, and handoffs. But until Brainclaw supports dedicated Git worktrees per agent/session, parallel edits in the same checkout are still likely to create more Git and workspace problems than they solve.
10
-
11
- For now, prefer:
12
-
13
- 1. one active editing agent per checkout
14
- 2. explicit handoffs between agents
15
- 3. claims and context to keep continuity between sessions
16
-
17
- ## 1. Bootstrap and initialize the workspace
3
+ This guide is organized by entry path, because Brainclaw serves different surfaces with different roles.
4
+
5
+ Use this rule first:
6
+
7
+ - capable agent with MCP support: prefer MCP for dynamic state
8
+ - agent surface driven mainly by local instruction files: use generated native files plus CLI fallback
9
+ - human operator or maintainer: use the CLI directly
10
+
11
+ ## Important limitation for now
12
+
13
+ Do not run multiple coding agents in parallel on the same project checkout yet.
14
+
15
+ Brainclaw is already useful for sequential collaboration: one agent can pick up where another stopped, inspect shared context, and continue from explicit plans, claims, traps, and handoffs. But until Brainclaw supports dedicated Git worktrees per agent/session, parallel edits in the same checkout are still likely to create more Git and workspace problems than they solve.
16
+
17
+ For now, prefer:
18
+
19
+ 1. one active editing agent per checkout
20
+ 2. explicit handoffs between agents
21
+ 3. claims and context to keep continuity between sessions
22
+
23
+ ## Path 1: Agent-First With MCP
24
+
25
+ Use this path when the agent can call Brainclaw through MCP.
26
+
27
+ ### Operator bootstrap
18
28
 
19
29
  ```bash
20
30
  brainclaw setup --yes
21
31
  brainclaw init
22
32
  ```
23
33
 
24
- `setup` installs the machine-level prerequisites and agent integrations. `init` then creates the workspace state, seeds stable identity, and prepares the project memory structure.
25
- If an AI coding agent is detected in the environment, brainclaw also writes to its native instruction file automatically.
34
+ `setup` installs machine-level prerequisites and agent integrations. `init` creates the workspace state, seeds stable identity, and prepares the project memory structure.
26
35
 
27
- ## 2. Capture the first important facts
36
+ ### Agent runtime pattern
28
37
 
29
- ```bash
30
- brainclaw memory create decision "OAuth migration now goes through auth-gateway" --tag auth
31
- brainclaw memory create constraint "Payments module frozen until 2026-04-01" --tag payments
32
- brainclaw memory create trap "Checkout E2E tests are flaky on Windows" --severity high --tag tests
38
+ After the workspace is initialized, the nominal flow is:
39
+
40
+ ```text
41
+ bclaw_session_start -> open a session and return current board/context
42
+ bclaw_get_context -> fetch fresh prompt-ready context for the target path
43
+ bclaw_list_plans -> inspect active work
44
+ bclaw_claim -> claim scope before editing
45
+ bclaw_write_note -> record runtime observations
46
+ bclaw_session_end -> close session cleanly and hand work off
33
47
  ```
34
48
 
35
- These become part of the shared project memory.
49
+ Use native agent files such as `AGENTS.md`, `CLAUDE.md`, or Cursor rules as local workflow guidance, not as the only source of live state.
50
+
51
+ ## Path 2: CLI-Oriented Agent Or Fallback Workflow
36
52
 
37
- ## 3. Create a shared plan
53
+ Use this path when the agent does not have a good MCP integration yet, or when a human needs to drive the workflow directly.
54
+
55
+ ### Bootstrap and inspect
38
56
 
39
57
  ```bash
40
- brainclaw plan create "Coordinate auth rollout" --priority high
41
- brainclaw plan list
58
+ brainclaw setup --yes
59
+ brainclaw init
60
+ brainclaw export --detect --write
42
61
  ```
43
62
 
44
- ## 4. Claim work before editing
63
+ ### Record the first important facts
45
64
 
46
65
  ```bash
47
- brainclaw claim create "Taking auth rollout" --agent copilot --scope src/auth/ --plan pln_001
48
- brainclaw claim list --plan pln_001
66
+ brainclaw memory create decision "OAuth migration now goes through auth-gateway" --tag auth
67
+ brainclaw memory create constraint "Payments module frozen until 2026-04-01" --tag payments
68
+ brainclaw memory create trap "Checkout E2E tests are flaky on Windows" --severity high --tag tests
49
69
  ```
50
70
 
51
- Claims reduce collisions, but they are not a substitute for isolated worktrees yet.
52
- Use them mainly to coordinate sequential work or human/agent awareness in the same repo.
53
-
54
- ## 5. Create an explicit handoff
71
+ ### Create and claim work
55
72
 
56
73
  ```bash
57
- brainclaw memory create handoff "Review auth patch" --from copilot --to claude --plan pln_001
74
+ brainclaw plan create "Coordinate auth rollout" --priority high
75
+ brainclaw claim create "Take auth rollout" --scope src/auth/
58
76
  ```
59
77
 
60
- ## 6. Generate context for an agent
78
+ ### Refresh context before edits
61
79
 
62
80
  ```bash
63
81
  brainclaw context --for src/auth/routes.ts --digest
64
- brainclaw context --json --max-chars 1200
82
+ brainclaw status
65
83
  ```
66
84
 
67
- Use this to prepare compact context before edits or reviews.
85
+ Claims reduce collisions, but they are not a substitute for isolated worktrees yet. Use them mainly to coordinate sequential work or human/agent awareness in the same repo.
86
+
87
+ ## Path 3: Brownfield Onboarding
68
88
 
69
- ## 7. Export to your agent's native instruction file
89
+ Use this path when you are adopting Brainclaw into an existing workspace and do not want to hand-author all memory from scratch.
90
+
91
+ ### Build the initial bootstrap view
70
92
 
71
93
  ```bash
72
- brainclaw export --detect # auto-detects running agent, writes to its file
73
- brainclaw export --format claude-md --write # writes CLAUDE.md and gitignores it by default
74
- brainclaw export --format cursor-rules --write # writes .cursor/rules/brainclaw.md and gitignores it by default
75
- brainclaw export --format claude-md --write --shared # only if you intentionally want to commit it
94
+ brainclaw setup --yes
95
+ brainclaw init
96
+ brainclaw bootstrap --json
76
97
  ```
77
98
 
78
- ## 8. Inspect the current board
99
+ ### Fill the gaps
79
100
 
80
101
  ```bash
81
- brainclaw status
82
- brainclaw agent-board --agent copilot
102
+ brainclaw bootstrap --interview --audience cli
103
+ brainclaw bootstrap --interview --audience ide_chat
83
104
  ```
84
105
 
85
- ## Recommended first workflow
106
+ ### Apply or rollback managed imports
86
107
 
87
- 1. initialize the workspace
88
- 2. record 3–5 important decisions or traps
89
- 3. create one shared plan
90
- 4. use claims for touched folders
91
- 5. generate context before edits
92
- 6. hand off explicitly when switching between agents
108
+ ```bash
109
+ brainclaw bootstrap --apply
110
+ brainclaw bootstrap --uninstall
111
+ ```
93
112
 
94
- ## Next reads
113
+ Use this path when the repo already has native instruction files, partial docs, or conventions that Brainclaw should adopt selectively instead of replacing blindly.
95
114
 
96
- - [cli.md](cli.md) full command reference
115
+ ## Recommended First Workflow
116
+
117
+ 1. initialize the workspace
118
+ 2. choose the correct entry path for your surface
119
+ 3. record or import 3-5 high-signal facts
120
+ 4. create one shared plan
121
+ 5. claim scope before editing
122
+ 6. refresh context before significant edits
123
+ 7. hand off explicitly when switching between agents
124
+
125
+ ## Next Reads
126
+
127
+ - [integrations/overview.md](integrations/overview.md) — integration model by surface
128
+ - [integrations/mcp.md](integrations/mcp.md) — nominal dynamic path for capable agents
129
+ - [cli.md](cli.md) — operator and fallback reference
97
130
  - [concepts/memory.md](concepts/memory.md)
98
131
  - [concepts/plans-and-claims.md](concepts/plans-and-claims.md)
99
- - [integrations/overview.md](integrations/overview.md)
package/package.json CHANGED
@@ -1,18 +1,18 @@
1
1
  {
2
2
  "name": "brainclaw",
3
- "version": "0.19.7",
3
+ "version": "0.19.10",
4
4
  "description": "Shared project memory for humans and coding agents.",
5
5
  "type": "module",
6
6
  "bin": {
7
7
  "brainclaw": "dist/cli.js",
8
8
  "bclaw": "dist/cli.js"
9
9
  },
10
- "files": [
11
- "dist/**/*.js",
12
- "docs/**/*.md",
13
- "README.md",
14
- "LICENSE"
15
- ],
10
+ "files": [
11
+ "dist/**/*.js",
12
+ "docs/**/*.md",
13
+ "README.md",
14
+ "LICENSE"
15
+ ],
16
16
  "scripts": {
17
17
  "build": "tsc",
18
18
  "dev": "tsc && node dist/cli.js",