@first-tree-ai/context-tree 0.1.7-alpha.202609010710 → 0.1.7
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 +35 -38
- package/dist/cli/index.mjs +923 -647
- package/package.json +7 -28
- package/scripts/postinstall.mjs +53 -0
- package/skills/context-tree-connect/SKILL.md +9 -7
- package/skills/context-tree-create/SKILL.md +10 -10
- package/skills/context-tree-publish/SKILL.md +4 -8
- package/skills/context-tree-read/SKILL.md +32 -12
- package/skills/context-tree-setup/SKILL.md +7 -12
- package/skills/context-tree-write/SKILL.md +196 -17
- package/.agents/plugins/marketplace.json +0 -19
- package/.claude-plugin/marketplace.json +0 -21
- package/.claude-plugin/plugin.json +0 -16
- package/.codex-plugin/plugin.json +0 -36
- package/dist/cli/index.d.mts +0 -1
- package/dist/index.d.mts +0 -77
- package/dist/index.mjs +0 -1449
- package/dist/schemas-C_7izpsa.d.mts +0 -390
- package/dist/schemas-DKHE1sWt.mjs +0 -289
- package/dist/schemas.d.mts +0 -2
- package/dist/schemas.mjs +0 -2
- package/docs/specification.md +0 -160
- package/examples/basic/NODE.md +0 -15
- package/examples/basic/members/NODE.md +0 -8
- package/examples/basic/members/example-agent/NODE.md +0 -7
- package/examples/basic/members/example-agent/memory.md +0 -9
- package/examples/basic/systems/NODE.md +0 -10
- package/examples/basic/systems/runtime.md +0 -15
- package/hooks/hooks.json +0 -26
- package/hooks/session-start.mjs +0 -55
- package/policy/context-tree-policy.md +0 -156
- package/skills/context-tree-connect/scripts/context-tree.mjs +0 -41
- package/skills/context-tree-create/scripts/context-tree.mjs +0 -41
- package/skills/context-tree-publish/scripts/context-tree.mjs +0 -41
- package/skills/context-tree-read/scripts/context-tree.mjs +0 -41
- package/skills/context-tree-setup/scripts/context-tree.mjs +0 -41
- package/skills/context-tree-write/scripts/context-tree.mjs +0 -41
package/README.md
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
# Context Tree
|
|
2
2
|
|
|
3
3
|
`@first-tree-ai/context-tree` provides durable, structured project context for
|
|
4
|
-
coding agents. It ships a portable core, CLI,
|
|
4
|
+
coding agents. It ships a portable core, CLI, templates, and six
|
|
5
5
|
framework-neutral skills.
|
|
6
6
|
|
|
7
7
|
A Context Tree records current decisions, constraints, relationships, and their
|
|
@@ -19,44 +19,42 @@ are credential-free `OWNER/REPO` identities, never URLs containing credentials.
|
|
|
19
19
|
|
|
20
20
|
## Install
|
|
21
21
|
|
|
22
|
-
### Codex or Claude Code plugin (recommended)
|
|
23
|
-
|
|
24
|
-
Install the marketplace and plugin, then start a new session so the host can
|
|
25
|
-
discover the skills and lifecycle hook.
|
|
26
|
-
|
|
27
|
-
For Codex:
|
|
28
|
-
|
|
29
22
|
```bash
|
|
30
|
-
|
|
31
|
-
codex plugin add context-tree@context-tree
|
|
23
|
+
npm install --global @first-tree-ai/context-tree
|
|
32
24
|
```
|
|
33
25
|
|
|
34
|
-
|
|
26
|
+
That installs the `context-tree` command and copies the six skills into the
|
|
27
|
+
skill directory of every agent you already have:
|
|
35
28
|
|
|
36
|
-
```
|
|
37
|
-
claude
|
|
38
|
-
|
|
29
|
+
```text
|
|
30
|
+
✓ claude → ~/.claude/skills/ (6 skills)
|
|
31
|
+
✓ codex → ~/.codex/skills/ (6 skills)
|
|
39
32
|
```
|
|
40
33
|
|
|
41
|
-
|
|
42
|
-
its private packaged CLI rather than a global `PATH` command — so plugin users
|
|
43
|
-
need no separate CLI installation. Review and trust the session-start hook if
|
|
44
|
-
your host asks. Then try asking:
|
|
34
|
+
Restart your agent so it discovers them, then try asking:
|
|
45
35
|
|
|
46
36
|
> Set up a Context Tree for this project, then read the relevant context.
|
|
47
37
|
|
|
48
38
|
> Write this architectural decision to the Context Tree.
|
|
49
39
|
|
|
50
|
-
|
|
51
|
-
|
|
52
|
-
Install globally only when scripts or terminal workflows need a `context-tree`
|
|
53
|
-
command on `PATH`:
|
|
40
|
+
Skill installation is a normal command, so you can re-run it after installing a
|
|
41
|
+
new agent, or scope it to one project:
|
|
54
42
|
|
|
55
43
|
```bash
|
|
56
|
-
|
|
57
|
-
context-tree --
|
|
44
|
+
context-tree install # every agent you have
|
|
45
|
+
context-tree install --host codex # one agent
|
|
46
|
+
context-tree install --project . # ./.claude/skills and ./.codex/skills
|
|
58
47
|
```
|
|
59
48
|
|
|
49
|
+
Installing only ever writes `context-tree-*` skill directories, never touches
|
|
50
|
+
skills it does not own, and never creates a configuration directory for an agent
|
|
51
|
+
that is not present. Adding support for another agent is one entry in the host
|
|
52
|
+
table in `src/core/install.ts`.
|
|
53
|
+
|
|
54
|
+
Once a project is connected, `create` and `connect` record the tree in the
|
|
55
|
+
project's own `AGENTS.md`, so any agent that reads instruction files knows the
|
|
56
|
+
tree exists without host-specific configuration.
|
|
57
|
+
|
|
60
58
|
## Six skills
|
|
61
59
|
|
|
62
60
|
### Setup
|
|
@@ -65,8 +63,7 @@ context-tree --help
|
|
|
65
63
|
connection. It asks whether to create a new tree or connect an existing one,
|
|
66
64
|
then delegates to the create or connect workflow rather than duplicating
|
|
67
65
|
lifecycle policy. Read and write invoke setup when the current project has no
|
|
68
|
-
connection
|
|
69
|
-
explicit confirmation.
|
|
66
|
+
connection. Setup never publishes without explicit confirmation.
|
|
70
67
|
|
|
71
68
|
### Create
|
|
72
69
|
|
|
@@ -142,9 +139,12 @@ preserves the worktree. Prepare again and reapply the intended semantic change
|
|
|
142
139
|
once; there is no automatic rebase, retry loop, or pull-request fallback.
|
|
143
140
|
|
|
144
141
|
A preserved or abandoned write leaves its temporary worktree on disk and a
|
|
145
|
-
`context-tree/write/<name>` branch in the tree.
|
|
146
|
-
|
|
147
|
-
|
|
142
|
+
`context-tree/write/<name>` branch in the tree. The next `prepare-write` reclaims
|
|
143
|
+
one of these only when it holds no commit your checkout lacks, has no pending
|
|
144
|
+
change, and has gone untouched for twenty-four hours, so a worktree you are still
|
|
145
|
+
editing and a `WRITE_OUTDATED` worktree awaiting its retry are both left alone.
|
|
146
|
+
Those keep their pending edits until you clear them with
|
|
147
|
+
`git worktree remove <path>` and `git branch -D <branch>` in the connected tree.
|
|
148
148
|
|
|
149
149
|
### Publish
|
|
150
150
|
|
|
@@ -181,15 +181,15 @@ commit or discard them), `INVALID_TREE` (structure fails `verify`),
|
|
|
181
181
|
The public command inventory is:
|
|
182
182
|
|
|
183
183
|
```text
|
|
184
|
-
create connect list resolve sync prepare-write
|
|
185
|
-
publish read verify
|
|
184
|
+
install create connect list resolve sync prepare-write
|
|
185
|
+
finish-write publish read verify
|
|
186
186
|
```
|
|
187
187
|
|
|
188
188
|
Setup, create, connect, read, write, and publish ship as six skills; setup
|
|
189
|
-
orchestrates the five concrete workflows. `
|
|
190
|
-
|
|
191
|
-
|
|
192
|
-
connect-target discovery.
|
|
189
|
+
orchestrates the five concrete workflows. `install` is the distribution
|
|
190
|
+
entry point, run for you by `npm install`. `resolve`, `sync`, `prepare-write`,
|
|
191
|
+
`finish-write`, and `verify` are plumbing or diagnostic commands rather than
|
|
192
|
+
separate user intentions; `list` backs setup's connect-target discovery.
|
|
193
193
|
All machine-readable responses use strict schema version `1`.
|
|
194
194
|
|
|
195
195
|
`verify` is intended for CI and diagnostics. Normal skills invoke it only after
|
|
@@ -202,10 +202,7 @@ pnpm install
|
|
|
202
202
|
pnpm check
|
|
203
203
|
pnpm typecheck
|
|
204
204
|
pnpm test
|
|
205
|
-
pnpm build
|
|
206
|
-
pnpm validate:skills
|
|
207
205
|
pnpm check:package
|
|
208
|
-
npm pack --dry-run
|
|
209
206
|
```
|
|
210
207
|
|
|
211
208
|
See [docs/specification.md](docs/specification.md) for contracts and safety
|