@atollhq/skill-claude 0.4.36 → 0.4.37

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/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@atollhq/skill-claude",
3
- "version": "0.4.36",
3
+ "version": "0.4.37",
4
4
  "description": "Install the Atoll project management skill for Claude Code",
5
5
  "bin": {
6
6
  "skill-claude": "bin/install.mjs"
@@ -2,6 +2,105 @@
2
2
 
3
3
  Read this reference before installing, diagnosing, configuring, or operating `atoll-runner`, its repository bindings, loopback UI, leases, or recovery behavior.
4
4
 
5
+ ## End-to-end setup
6
+
7
+ The released runner does not require an Atoll source checkout. It is the
8
+ `atoll-runner` binary in `@atollhq/cli`; the machine needs Node.js, Git, a local
9
+ checkout of each repository that will receive work, and an authenticated Codex
10
+ runtime. On macOS, first confirm that `git --version` succeeds. Apple Git can be
11
+ blocked until the operator reviews and accepts the Xcode Command Line Tools
12
+ license; do not accept legal terms or enter an administrator password for them.
13
+
14
+ Install or update the released CLI, then verify that both binaries exist:
15
+
16
+ ```bash
17
+ npm install -g @atollhq/cli@latest
18
+ atoll --version
19
+ atoll-runner --help
20
+ ```
21
+
22
+ Create one named Atoll profile per agent identity and organization. Obtain the
23
+ agent key from the human operator; never print it, put it in shell history on
24
+ their behalf, or copy it into runner configuration. The interactive flow is
25
+ preferred when a terminal is available:
26
+
27
+ ```bash
28
+ atoll auth setup
29
+ atoll auth profiles --json
30
+ ```
31
+
32
+ For each profile, discover the project's verified repository mapping and keep
33
+ its opaque `repo_ref`. Do not invent a reference or treat a local path as
34
+ authorization:
35
+
36
+ ```bash
37
+ atoll --profile agent-a repository list --project project-slug --json
38
+ atoll-runner --profile agent-a repositories bind repo-ref /absolute/path/to/checkout --issue issue-uuid
39
+ atoll-runner --profile agent-a repositories validate repo-ref
40
+ ```
41
+
42
+ The issue is authorization evidence for its project; binding does not execute
43
+ that issue. The checkout must be a non-bare Git repository whose credential-free
44
+ `origin` matches the authorized GitHub repository. The runner uses the checkout
45
+ as a verified source and creates isolated worktrees under its private local
46
+ state. It does not work directly in the primary checkout, and the released CLI
47
+ does not accept a custom worktree root or per-issue working directory.
48
+
49
+ Before the first start, the authenticated agent has no hosted runner row to
50
+ pause. The operator must first remove every eligible assignment from that agent,
51
+ or use a dedicated setup agent with no assignments. Do not infer safety from a
52
+ quick issue read when another process can assign work concurrently. With this
53
+ no-work precondition established, run one bounded real cycle to initialize local
54
+ state and register the hosted row:
55
+
56
+ ```bash
57
+ atoll-runner --profile agent-a run --once
58
+ ```
59
+
60
+ This is not a model test. It is safe only because the agent has no eligible
61
+ assigned work. A binding command creates local identity but not runner state or
62
+ server presence; `doctor` will not fully pass yet, and `--dry-run` intentionally
63
+ does not perform registration. As soon as the runner appears, pause it in
64
+ **Workspace Settings → Runners → Manage → Pause new work** and verify that the
65
+ fleet row shows **Paused**. Keep it paused while running the read-only checks and
66
+ installing the optional macOS service:
67
+
68
+ ```bash
69
+ atoll-runner --profile agent-a doctor
70
+ atoll-runner --profile agent-a run --once --dry-run
71
+ atoll-runner --profile agent-a service install --ui --ui-port 4735
72
+ atoll-runner --profile agent-a service status
73
+ atoll-runner --profile agent-a status
74
+ ```
75
+
76
+ Expected readiness evidence is: `doctor` succeeds, every required repository
77
+ binding validates, Codex is authenticated, the service is installed and loaded,
78
+ the loopback UI returns HTTP 200 when enabled, and runner status reports Atoll
79
+ reachable with no unexpected current or retained job. A dry run is read-only;
80
+ after the hosted row is paused it should return no candidate with
81
+ `blocked_reason: intake_paused`.
82
+
83
+ Only after those checks pass should the operator resume **Workspace Settings →
84
+ Runners → Manage → Resume new work**. Read back both the hosted fleet row and
85
+ `atoll-runner --profile agent-a status`; the final state is `connected` with
86
+ active/accepting intake. Resuming can immediately claim eligible assigned work,
87
+ so do not use it as a setup probe.
88
+
89
+ One machine can store and run multiple named profiles, including profiles for
90
+ different organizations. Each runner process and macOS service is fixed to one
91
+ profile, and that profile supplies one agent identity and organization. Install
92
+ one service per profile and give every enabled loopback UI a different port:
93
+
94
+ ```bash
95
+ atoll-runner --profile agent-a service install --ui --ui-port 4735
96
+ atoll-runner --profile agent-b service install --ui --ui-port 4736
97
+ ```
98
+
99
+ Do not run competing profiles for the same Atoll agent identity: the server
100
+ allows one current local runner installation per authenticated agent. A single
101
+ profile can bind several authorized repositories, each to its own absolute local
102
+ checkout. Work remains serialized to one current job per runner installation.
103
+
5
104
  ### Local runner presence
6
105
 
7
106
  Authenticated agents can register and refresh one local runner installation with