@kaddo/mcp 3.19.0 → 3.20.0
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/README.md +30 -4
- package/dist/index.js +2561 -84
- package/package.json +2 -2
package/README.md
CHANGED
|
@@ -60,6 +60,32 @@ directory set via the `KADDO_PROJECT_DIR` environment variable).
|
|
|
60
60
|
- `kaddo_list_agents` / `kaddo_get_agent_prompt` — installed agent prompts.
|
|
61
61
|
- `kaddo_list_graph_hints` — graph hints, filter by `artifact_type` / `severity` / `active_only`.
|
|
62
62
|
|
|
63
|
+
## Derived tools (write only under `.kaddo/`)
|
|
64
|
+
|
|
65
|
+
When a derived artifact is missing or stale, these tools regenerate it in place using the same core
|
|
66
|
+
logic as the CLI — so the agent never has to leave the flow. They are deterministic (no LLM) and
|
|
67
|
+
**only write under `.kaddo/`**; they never touch `knowledge/`, `src/`, `external/`,
|
|
68
|
+
`.kaddo/external.yml` or git.
|
|
69
|
+
|
|
70
|
+
| Tool | Writes | CLI equivalent |
|
|
71
|
+
|---|---|---|
|
|
72
|
+
| `kaddo_generate_context` | `.kaddo/context-pack.md` + `.json` | `kaddo context` |
|
|
73
|
+
| `kaddo_generate_explain` | `.kaddo/explain.md` + `.json` | `kaddo explain` |
|
|
74
|
+
| `kaddo_generate_understand` | `.kaddo/understand.md` | `kaddo understand` |
|
|
75
|
+
| `kaddo_generate_graph` | `.kaddo/graph.json` + `.mmd` + `graph-hints.md` + `.json` | `kaddo graph export` |
|
|
76
|
+
| `kaddo_generate_capsule_draft` | `.kaddo/exports/<project>.capsule.md` + `.json` | `kaddo capsule export` |
|
|
77
|
+
|
|
78
|
+
`kaddo_generate_capsule_draft` writes a **draft only** — it never registers/imports a capsule
|
|
79
|
+
(`external/`, `.kaddo/external.yml` stay untouched; use the CLI `kaddo capsule add` for that).
|
|
80
|
+
|
|
81
|
+
Each tool returns `{ status, files_written, summary, warnings, next_suggested_resources }`. All
|
|
82
|
+
writes pass through a central allowlist (`assertMcpDerivedWritePath`) that blocks any path outside
|
|
83
|
+
the derived `.kaddo/` set — attempts return `Blocked unsafe MCP derived write path.`
|
|
84
|
+
|
|
85
|
+
**When to use:** after a resource reports a missing file (e.g. `Run kaddo context first`), call the
|
|
86
|
+
matching derived tool, then read the resource again. Whether a tool runs automatically or needs
|
|
87
|
+
confirmation is up to your MCP client.
|
|
88
|
+
|
|
63
89
|
## Prompts
|
|
64
90
|
|
|
65
91
|
Every installed agent prompt (`knowledge/agents/**`) is exposed as an MCP prompt
|
|
@@ -68,10 +94,10 @@ content and recommended inputs.
|
|
|
68
94
|
|
|
69
95
|
## What it does NOT do
|
|
70
96
|
|
|
71
|
-
No writes, no Work Item creation, no `kaddo scan
|
|
72
|
-
|
|
73
|
-
|
|
74
|
-
|
|
97
|
+
No writes outside `.kaddo/`, no knowledge/source edits, no Work Item creation, no `kaddo scan`/
|
|
98
|
+
`learn`/`owners suggest`/`capsule add`, no `kaddo add`, no git, no remote sync, no GitHub API, no
|
|
99
|
+
HTTP server, no auth, no RAG, no vector database, no LLM calls. The derived tools above may
|
|
100
|
+
regenerate artifacts **only under `.kaddo/`**; everything else is read-only.
|
|
75
101
|
|
|
76
102
|
## Security
|
|
77
103
|
|