@rse/ase 0.9.62 → 0.9.63

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.
@@ -20,20 +20,23 @@ kinds (the *source*). It reads the source artifacts and then adjusts the
20
20
  target artifacts *directly* and *surgically* to match the source state,
21
21
  while optionally honoring a filtering *hint*.
22
22
 
23
- Both *target* and *source* are comma-separated lists over the seven
24
- recognized artifact kinds `SPEC` (Specification), `ARCH` (Architecture),
25
- `CODE` (Source Code), `DOCS` (Documentation), `TASK` (Task Plans), `INFR`
26
- (Infrastructure), and `OTHR` (catch-all). When *source* is `auto`, it
27
- resolves to all seven kinds *minus* the kinds listed in *target*. Unless
28
- `--bidirectional` is given, a kind present in *target* is never used as
29
- its own source.
23
+ Both *target* and *source* are comma-separated lists over the six
24
+ recognized artifact kinds `SPEC` (Specification, covering both
25
+ requirements and architecture), `CODE` (Source Code), `DOCS`
26
+ (Documentation), `TASK` (Task Plans), `INFR` (Infrastructure), and
27
+ `OTHR` (catch-all). When *source* is `auto`, it resolves to all six
28
+ kinds *minus* the kinds listed in *target*. Unless `--bidirectional` is
29
+ given, a kind present in *target* is never used as its own source.
30
30
 
31
31
  The file lists for all involved kinds are resolved via the
32
32
  `ase_artifact_list` MCP tool of the `ase` MCP server. While reconciling,
33
33
  the skill honors the artifact-format conventions of `ase-format-meta.md`,
34
- `ase-format-spec.md`, `ase-format-arch.md`, and `ase-format-task.md`;
35
- the kinds `CODE`, `DOCS`, `INFR`, and `OTHR` have no dedicated format
36
- contract and are treated as free-form.
34
+ `ase-format-spec.md` (the *SpecBook* models and formats plus the
35
+ SpecBook schema configuration), and `ase-format-task.md`; the kinds
36
+ `CODE`, `DOCS`, `INFR`, and `OTHR` have no dedicated format contract and
37
+ are treated as free-form. Changed `SPEC` artifacts are validated via
38
+ the `ase_specbook_lint` MCP tool and the reported diagnostics are fixed
39
+ in at most three rounds; any remaining diagnostics are surfaced.
37
40
 
38
41
  ## OPTIONS
39
42
 
@@ -62,25 +65,25 @@ contract and are treated as free-form.
62
65
  ## EXAMPLES
63
66
 
64
67
  Reconcile the code and documentation
65
- to reflect the current specification and architecture
68
+ to reflect the current specification
66
69
  in a "forward engineering" approach:
67
70
 
68
71
  ```text
69
- ❯ /ase-sync-reconcile -t CODE,DOCS -s SPEC,ARCH
72
+ ❯ /ase-sync-reconcile -t CODE,DOCS -s SPEC
70
73
  ```
71
74
 
72
- Reconcile specification and architecture from everything else
75
+ Reconcile the specification from everything else
73
76
  in a "reverse engineering" approach:
74
77
 
75
78
  ```text
76
- ❯ /ase-sync-reconcile -t SPEC,ARCH -s CODE,DOCS
79
+ ❯ /ase-sync-reconcile -t SPEC -s CODE,DOCS
77
80
  ```
78
81
 
79
- Bidirectionally reconcile specification and architecture against
82
+ Bidirectionally reconcile specification and code against
80
83
  each other, limited to the authentication aspect:
81
84
 
82
85
  ```text
83
- ❯ /ase-sync-reconcile -b -t SPEC -s ARCH authentication
86
+ ❯ /ase-sync-reconcile -b -t SPEC -s CODE authentication
84
87
  ```
85
88
 
86
89
  ## SEE ALSO
@@ -141,10 +141,11 @@ Set <args>--int-reuse-task</args>.
141
141
  `ARCHITECTURE`, and then all `IMPLEMENTATION` ones -- and
142
142
  *secondarily*, within each focus area, by the decision tree
143
143
  order, so that each decision is asked *after* the decisions
144
- it depends on. Renumber <N/> according to this order.
145
- Truncate the list after a maximum of 10 questions and set
146
- <n/> to the number of remaining questions. Do not output
147
- anything.
144
+ it depends on. Renumber <N/> according to this order,
145
+ starting at `1` in *every* round, independent of the
146
+ numbering of previous rounds. Truncate the list after a
147
+ maximum of 10 questions and set <n/> to the number of
148
+ remaining questions. Do not output anything.
148
149
 
149
150
  5. For each question <question-N/> in the iteration cycle <N/>,
150
151
  *one at a time*: