@jarkkojs/goosedump 0.12.31 → 0.12.32

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.
Files changed (2) hide show
  1. package/README.md +76 -27
  2. package/package.json +5 -5
package/README.md CHANGED
@@ -1,29 +1,30 @@
1
1
  # goosedump
2
2
 
3
3
  `goosedump` browses, searches, converts, and compacts saved coding-agent
4
- sessions. It also provides project-scoped durable memory for Claude Code,
5
- Codex, Crush, Gemini CLI, Goose, OpenCode, and Pi.
4
+ sessions. It also provides project-scoped durable memory.
5
+
6
+ Supported agents: Claude Code, Codex, Crush, Gemini CLI, Goose, OpenCode, and Pi.
6
7
 
7
8
  It builds on ideas from [`lllyasviel/VCC`](https://github.com/lllyasviel/VCC).
8
9
 
9
10
  ## Install
10
11
 
11
- Prebuilt binaries support Linux x64 and arm64, macOS arm64, and Windows x64:
12
+ Prebuilt binaries support Linux x64/arm64, macOS arm64, and Windows x64:
12
13
 
13
14
  ```sh
14
15
  npm install --global @jarkkojs/goosedump
15
16
  ```
16
17
 
17
- To build from source:
18
+ Build from source (Rust 1.95+, C compiler, Node.js 18+):
18
19
 
19
20
  ```sh
20
21
  cargo build --release
21
22
  ```
22
23
 
23
- Requires Node.js 18 or later. Source builds require Rust 1.95 and a C compiler.
24
-
25
24
  ## Quick start
26
25
 
26
+ Session targets use `provider:session-id`. Bare IDs are not accepted.
27
+
27
28
  ```sh
28
29
  goosedump session list --query '.provider == "pi"'
29
30
  goosedump session show pi:context-id
@@ -34,34 +35,80 @@ goosedump memory learn pi:context-id
34
35
  goosedump memory recall 'database failure'
35
36
  ```
36
37
 
37
- Commands that address a saved session require a `provider:session-id` target.
38
- See `goosedump --help` and the [manual page](man/man1/goosedump.1) for options.
38
+ Full options: `goosedump --help` and the [manual page](man/man1/goosedump.1).
39
+
40
+ ## Commands
41
+
42
+ | Group | Commands |
43
+ | --------- | ---------------------------------------------------------------- |
44
+ | `session` | `list`, `show`, `find`, `search`, `compact`, `copy`, `remove` |
45
+ | `memory` | `learn`, `recall`, `list`, `show`, `forget`, `status` |
46
+
47
+ ### Session
48
+
49
+ | Command | Purpose |
50
+ | --------- | ---------------------------------------------------------------- |
51
+ | `list` | Discover sessions as `provider:id` targets |
52
+ | `show` | Print a transcript (optionally re-render with `--as`) |
53
+ | `find` | Glob-match message text |
54
+ | `search` | Rank messages by relevance (5 results per page) |
55
+ | `compact` | Create a summary; does not write durable memory |
56
+ | `copy` | Preview or write a cross-provider copy |
57
+ | `remove` | Preview or remove a session and its descendants |
58
+
59
+ Range filters (`--entry`, `--from`, `--before`, `--all`) apply to `show`,
60
+ `find`, `search`, and `compact`.
39
61
 
40
- ## Compaction and memory
62
+ ### Memory
41
63
 
42
- Compaction creates a summary and does not change durable memory. Summaries have
43
- an estimated 4,096-token limit by default; pass `--summary-max-tokens 0` to
44
- disable it.
64
+ Compaction and memory are separate. `memory learn` stores evidence first, then
65
+ derives cited atomic claims. Recall searches the current project by default.
45
66
 
46
- `memory learn` stores evidence first, then derives cited atomic claims from new
47
- session entries. Recall searches the current project by default, combining
48
- lexical, entity, relation, and semantic matches. Use `--history` for superseded
49
- claims or `--all-projects` to search every project.
50
- Recall JSON includes `match_reasons` and cited evidence records with stable
51
- `citation_id` values and snippets.
67
+ | Command | Purpose |
68
+ | -------- | ----------------------------------------------------------------- |
69
+ | `learn` | Learn from new entries in a session's active lineage |
70
+ | `recall` | Search claims (lexical, entity, relation, then semantic) |
71
+ | `list` | List recent claims |
72
+ | `show` | Show one claim with cited evidence |
73
+ | `forget` | Tombstone a claim or a session's retained evidence |
74
+ | `status` | Storage, indexing, embedding, and type counts |
52
75
 
53
- The text and embedding models run locally and download on first use. Evidence,
54
- claims, entities, relations, and vectors are stored together in SQLite.
76
+ Memory types: `decision`, `fact`, `preference`, `procedure`, `lesson`.
77
+
78
+ Defaults:
79
+
80
+ | Setting | Default |
81
+ | ----------------------- | ------: |
82
+ | Compact summary budget | 4096 |
83
+ | Recall result limit | 10 |
84
+ | Recall token budget | 2048 |
85
+
86
+ Pass `--summary-max-tokens 0` to disable the compact limit. Use `--history` for
87
+ superseded claims and `--all-projects` to search every project. Recall JSON
88
+ includes `match_reasons` and cited evidence with stable `citation_id` values.
89
+
90
+ Text and embedding models run locally and download on first use. Evidence,
91
+ claims, entities, relations, and vectors live together in SQLite.
55
92
 
56
93
  ## Pi extension
57
94
 
58
- The bundled [pi-goosedump extension](packages/pi-goosedump/README.md) adds
59
- session search, memory, session management, and compaction to Pi:
95
+ The [pi-goosedump](packages/pi-goosedump/README.md) extension adds session
96
+ search, memory, session management, and compaction to Pi:
60
97
 
61
98
  ```sh
62
99
  pi install npm:pi-goosedump
63
100
  ```
64
101
 
102
+ ## Local models
103
+
104
+ | Model | Role |
105
+ | ----------------------------------------------------- | ------------------------------------------ |
106
+ | [gpt-oss-20b](docs/models/gpt-oss-20b.md) | Claim extraction; compact candidate filter |
107
+ | [bge-small-en-v1.5](docs/models/bge-small-en-v1.5.md) | Semantic claim retrieval |
108
+
109
+ Benchmarks for the text model:
110
+ [gpt-oss-20b-benchmarks](docs/models/gpt-oss-20b-benchmarks.md).
111
+
65
112
  ## Development
66
113
 
67
114
  ```sh
@@ -72,9 +119,11 @@ cargo fmt --check
72
119
 
73
120
  ## License
74
121
 
75
- - `goosedump`: LGPL-2.1-or-later
76
- - `pi-goosedump`: Apache-2.0
77
- - GPT-OSS-20B: Apache-2.0
78
- - bge-small-en-v1.5: MIT
122
+ | Component | License |
123
+ | ----------------- | ----------------- |
124
+ | goosedump | LGPL-2.1-or-later |
125
+ | pi-goosedump | Apache-2.0 |
126
+ | GPT-OSS-20B | Apache-2.0 |
127
+ | bge-small-en-v1.5 | MIT |
79
128
 
80
- Models are downloaded at run time and are not distributed with goosedump.
129
+ Models download at run time and are not distributed with goosedump.
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@jarkkojs/goosedump",
3
- "version": "0.12.31",
3
+ "version": "0.12.32",
4
4
  "description": "Coding agent context data browser",
5
5
  "license": "Apache-2.0 AND LGPL-2.1-or-later",
6
6
  "homepage": "https://github.com/jarkkojs/goosedump#readme",
@@ -21,10 +21,10 @@
21
21
  "LICENSE-LGPL-2.1"
22
22
  ],
23
23
  "optionalDependencies": {
24
- "@jarkkojs/goosedump-darwin-arm64": "0.12.31",
25
- "@jarkkojs/goosedump-linux-arm64": "0.12.31",
26
- "@jarkkojs/goosedump-linux-x64": "0.12.31",
27
- "@jarkkojs/goosedump-win32-x64": "0.12.31"
24
+ "@jarkkojs/goosedump-darwin-arm64": "0.12.32",
25
+ "@jarkkojs/goosedump-linux-arm64": "0.12.32",
26
+ "@jarkkojs/goosedump-linux-x64": "0.12.32",
27
+ "@jarkkojs/goosedump-win32-x64": "0.12.32"
28
28
  },
29
29
  "publishConfig": {
30
30
  "access": "public"