@skill-map/cli 0.53.3 → 0.53.5

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.
@@ -32,9 +32,9 @@ Wait for confirmation. Mark `init`: done.
32
32
 
33
33
  ## Chapter `kinds` - The other kinds appear (~1 min)
34
34
 
35
- Leave the browser open and the terminal with `sm` running. You create four more nodes **without any cross-fixture links** yet, pure standalone nodes, so the tester sees four new dots pop in. Three new **kinds** show up in this step (skill, command, markdown), the fourth file is a second `markdown` node that the hub in the `connectors` chapter will point at via a real `references` link.
35
+ Leave the browser open and the terminal with `sm` running. You create five more nodes **without any cross-fixture links** yet, pure standalone nodes, so the tester sees five new dots pop in. Three new **kinds** show up in this step (skill, command, markdown); the last two files are sibling `markdown` notes (`demo-guideline`, `demo-guideline2`) the hub in the `connectors` chapter reaches two ways, a faint mention that can't resolve and the same handle plus `.md` that resolves to a real file.
36
36
 
37
- Create these four files (with `Write`), exactly in this order. Per §Provider detection, **substitute `.claude/` with the detected `<provider_dir>` and skip files whose kind is not in the provider's supported set** (`agent-skills` / Antigravity: skip both `demo-agent` and `demo-command`, only the skill + the two markdown notes remain). Adjust the node count, the "four new nodes" message, and the file list shown to the tester in the sample below accordingly:
37
+ Create these five files (with `Write`), exactly in this order. Per §Provider detection, **substitute `.claude/` with the detected `<provider_dir>` and skip files whose kind is not in the provider's supported set** (`agent-skills` / Antigravity: skip both `demo-agent` and `demo-command`, only the skill + the three markdown notes remain). Adjust the node count, the "five new nodes" message, and the file list shown to the tester in the sample below accordingly:
38
38
 
39
39
  1. `.claude/skills/demo-skill/SKILL.md` (kind: skill):
40
40
  ```markdown
@@ -102,16 +102,16 @@ Create these four files (with `Write`), exactly in this order. Per §Provider de
102
102
  # Pending
103
103
  ```
104
104
 
105
- 4. `notes/demo-guideline.md`, second `kind: markdown` node, the
106
- one the hub will reach via a real markdown link in the
107
- `connectors` chapter:
105
+ 4. `notes/demo-guideline.md`, second `kind: markdown` node, reached
106
+ in the `connectors` chapter by a bare `@`-mention that can't
107
+ resolve, so it stays skill-map's faintest connector:
108
108
  ```markdown
109
109
  ---
110
110
  name: demo-guideline
111
111
  description: |
112
- Static reference notes that the rest of the demo points at.
113
- Showcases a second markdown node so the demo can exercise
114
- the `references` link kind without ambiguity.
112
+ Static reference notes the rest of the demo points at. The hub
113
+ reaches it with a bare `@`-mention, which stays a faint guess
114
+ (0.50) because it can't resolve to a known entity.
115
115
  tags: [notes, demo]
116
116
  ---
117
117
 
@@ -124,27 +124,49 @@ Create these four files (with `Write`), exactly in this order. Per §Provider de
124
124
  - Body stays minimal, only what's needed to teach the kind.
125
125
  ```
126
126
 
127
+ 5. `notes/demo-guideline2.md`, a sibling `markdown` node identical
128
+ to `demo-guideline`, reached by the same handle plus a `.md`
129
+ extension (`@demo-guideline2.md`), which makes it a file reference
130
+ that resolves to this node and lands at full confidence:
131
+ ```markdown
132
+ ---
133
+ name: demo-guideline2
134
+ description: |
135
+ Sibling of demo-guideline. The hub reaches it with an
136
+ `@`-mention that carries the `.md` extension, so the link
137
+ resolves to this file and lands certain (1.00).
138
+ tags: [notes, demo]
139
+ ---
140
+
141
+ # Demo Guideline 2
142
+
143
+ Same conventions as demo-guideline; it exists so the hub can
144
+ reach it with a resolved reference instead of a bare mention.
145
+ ```
146
+
127
147
  Tell the tester:
128
148
 
129
- > Look at the browser. Four new nodes should have popped in:
130
- > `demo-skill`, `demo-command`, **Demo TODO list**, and `demo-guideline`.
131
- > Five total now, **still unconnected**: they're floating dots.
149
+ > Look at the browser. Five new nodes should have popped in:
150
+ > `demo-skill`, `demo-command`, **Demo TODO list**, `demo-guideline`,
151
+ > and `demo-guideline2`.
152
+ > Six total now, **still unconnected**: they're floating dots.
132
153
  > The viewport auto-fits whenever a node is added or removed, so
133
- > all five should be visible without panning.
154
+ > all six should be visible without panning.
134
155
  >
135
- > What I just did behind the scenes: I created four new files in
156
+ > What I just did behind the scenes: I created five new files in
136
157
  > your project, and the watcher picked them up on its own, that's
137
- > why four new dots appeared without you running anything:
158
+ > why five new dots appeared without you running anything:
138
159
  >
139
160
  > - `.claude/skills/demo-skill/SKILL.md` (kind: skill)
140
161
  > - `.claude/commands/demo-command.md` (kind: command)
141
162
  > - `notes/todo.md` (kind: markdown)
142
163
  > - `notes/demo-guideline.md` (kind: markdown)
164
+ > - `notes/demo-guideline2.md` (kind: markdown)
143
165
  >
144
166
  > Same loop you'll use yourself in the next step, only this time
145
167
  > the writes came from me.
146
168
  >
147
- > Did the four appear? Confirm so we can wire them up.
169
+ > Did the five appear? Confirm so we can wire them up.
148
170
 
149
171
  Wait for confirmation. Mark `kinds`: done.
150
172
 
@@ -168,24 +190,24 @@ Tell the tester:
168
190
  >
169
191
  > Watch the browser. The `demo-agent` card should refresh its
170
192
  > description in real time, no reload, no Ctrl+C, same watcher
171
- > that picked up the four new nodes a moment ago, this time
193
+ > that picked up the five new nodes a moment ago, this time
172
194
  > reacting to YOUR edit.
173
195
  >
174
- > Confirm so we wire the five up.
196
+ > Confirm so we wire the six up.
175
197
 
176
198
  Wait for confirmation. You MAY use `Read` on the file afterwards to verify the change landed (read-only, allowed under Inviolable rule #1) before moving on. Mark `first-edit`: done.
177
199
 
178
200
  ## Chapter `connectors` - The connectors light up (~2 min)
179
201
 
180
- You edit `notes/todo.md` so it becomes the **hub** that points to each of the other four nodes. Each bullet uses a syntax that maps to a specific **link kind**:
202
+ You edit `notes/todo.md` so it becomes the **hub** that points to each of the other five nodes. Each bullet uses a syntax that maps to a specific **link kind**:
181
203
 
182
204
  - an `@handle` token → kind `mentions`
205
+ - an `@handle.md` token (a `@` handle that ends in a file extension) → kind `references`
183
206
  - a `/slash` token → kind `invokes`
184
- - a markdown link `[text](path)` → kind `references`
185
207
 
186
- Five bullets, three kinds: `invokes` and `mentions` each appear twice, `references` once. The last two bullets both point at the **same** node (`demo-guideline`), on purpose: a markdown link AND an `@`-mention of it, so the next beat can contrast their confidence (the file link is certain, the name drop is a softer guess).
208
+ Five bullets, three kinds: `invokes` and `mentions` each appear twice, `references` once. The last two bullets are the confidence lesson: a bare `@demo-guideline` mention (which can't resolve, so it stays a faint guess) next to `@demo-guideline2.md`, the same handle shape plus a `.md` extension that points at a real sibling file (so it resolves and lands certain). Two separate nodes, two clearly different confidences.
187
209
 
188
- Apply with `Edit` on `notes/todo.md` (do not rewrite the file). Per §Provider detection, **substitute `.claude/` with the detected `<provider_dir>` and drop any bullet whose target node was not created in the `kinds` chapter** (on `agent-skills` / Antigravity there is no agent and no command → skip the `@demo-agent` and `/demo-command` bullets; the `@demo-guideline` mention stays and is the faint connector on those providers too).
210
+ Apply with `Edit` on `notes/todo.md` (do not rewrite the file). Per §Provider detection, **substitute `.claude/` with the detected `<provider_dir>` and drop any bullet whose target node was not created in the `kinds` chapter** (on `agent-skills` / Antigravity there is no agent and no command → skip the `@demo-agent` and `/demo-command` bullets; the two guideline bullets stay, so the confidence contrast, faint mention 0.50 vs resolved reference 1.00, is intact on those providers too).
189
211
 
190
212
  **Edit `notes/todo.md`**: append these bullets after the `# Pending` heading:
191
213
 
@@ -193,9 +215,8 @@ Apply with `Edit` on `notes/todo.md` (do not rewrite the file). Per §Provider d
193
215
  - [ ] Brief @demo-agent on the rough edges.
194
216
  - [ ] Run /demo-command before publishing.
195
217
  - [ ] Trigger /demo-skill when the input lands.
196
- - [ ] Re-read the
197
- [demo-guideline](./demo-guideline.md) before shipping.
198
218
  - [ ] Ping @demo-guideline if the conventions change.
219
+ - [ ] Ping @demo-guideline2.md if the conventions change.
199
220
  ```
200
221
 
201
222
  Tell the tester:
@@ -207,22 +228,32 @@ Tell the tester:
207
228
  > - `Demo TODO list → demo-agent` (kind: `mentions`)
208
229
  > - `Demo TODO list → demo-command` (kind: `invokes`)
209
230
  > - `Demo TODO list → demo-skill` (kind: `invokes`)
210
- > - `Demo TODO list → demo-guideline` (kind: `references`, the markdown link)
211
- > - `Demo TODO list → demo-guideline` (kind: `mentions`, the @ handle)
231
+ > - `Demo TODO list → demo-guideline` (kind: `mentions`, the bare `@` handle)
232
+ > - `Demo TODO list → demo-guideline2` (kind: `references`, the `@` handle with a `.md` extension)
233
+ >
234
+ > The kind comes from the syntax in the bullet: an `@handle` is a
235
+ > mention, a `/skill` or `/command` is an invoke, and an `@handle`
236
+ > that ends in a file extension (`@name.md`) is a reference, the
237
+ > extension turns the name drop into a file pointer. Five arrows,
238
+ > three kinds.
239
+ >
240
+ > Now look closely: the two guideline arrows are not equally solid.
241
+ > Skill-map draws each connector's **confidence** as opacity, how
242
+ > sure it is the link resolves to something real:
212
243
  >
213
- > The kind comes from the syntax in the bullet: an `@handle` is
214
- > always a mention, a `/skill` or `/command` is always an invoke, a
215
- > markdown link is always a reference. Five arrows, three kinds.
244
+ > - `@demo-guideline` is a bare `@`-mention. A bare mention only
245
+ > firmly resolves to an *agent*, and `demo-guideline` is a note,
246
+ > not an agent, so skill-map keeps it a soft guess (0.50), drawn
247
+ > faint. (That's also why `@demo-agent` above is solid: it does
248
+ > resolve to a real agent.)
249
+ > - `@demo-guideline2.md` adds the `.md`, so skill-map reads a file
250
+ > pointer, finds the real `demo-guideline2.md` node, and is certain
251
+ > (1.00), drawn solid.
216
252
  >
217
- > Now look closely: the two arrows to **demo-guideline** are not
218
- > equally solid. Skill-map draws each connector's **confidence** as
219
- > opacity, how sure it is the link is real. The markdown link
220
- > `[demo-guideline](./demo-guideline.md)` points at a real file, so
221
- > it's certain (1.00) and looks solid; the `@demo-guideline` mention
222
- > is a looser name drop, a softer guess (0.50), so it's drawn fainter.
223
- > A glance at the map tells you which links are rock solid and which
224
- > are skill-map's best guess; the exact number per link is in the
225
- > inspector, next chapter.
253
+ > Same handle, one `.md` apart, and the confidence jumps from a guess
254
+ > to a certainty. A glance at the map tells you which links are rock
255
+ > solid and which are skill-map's best guess; the exact number per
256
+ > link is in the inspector, next chapter.
226
257
  >
227
258
  > Confirm when you see it. If a connector is missing, refresh the
228
259
  > browser and let me know.
@@ -238,12 +269,11 @@ The connector opacity tells the confidence story at a glance; the exact per-link
238
269
  > by default): it has two sections, **Outgoing** and **Incoming**.
239
270
  > Demo TODO list lists **5 links** under Outgoing (the hub) and 0
240
271
  > under Incoming; open the Inspector for a targeted node to see its
241
- > Incoming count (demo-guideline shows **2**, the reference plus the
242
- > mention; the others 1 each). Each row shows the link kind
243
- > (`mentions`, `invokes`, `references`) and a badge with its
244
- > confidence: the numeric value. Here you'll see the contrast in
245
- > numbers, the `references` to demo-guideline reads `1.00`, the
246
- > `mentions` of it reads `0.50`.
272
+ > Incoming count (each node the hub points at shows **1**). Each row
273
+ > shows the link kind (`mentions`, `invokes`, `references`) and a
274
+ > badge with its confidence: the numeric value. Here you'll see the
275
+ > contrast in numbers, the `mentions` to `demo-guideline` reads
276
+ > `0.50`, while the `references` to `demo-guideline2` reads `1.00`.
247
277
  >
248
278
  > 💡 Tip: if all these changes left the nodes crowded together, the
249
279
  > map toolbar has a **Re-arrange layout** button (tooltip "Re-arrange
@@ -283,7 +313,7 @@ Per §Provider detection, on `agent-skills` / Antigravity the fixture has fewer
283
313
  > by default: click the expand handle there (the `>` arrow, its
284
314
  > tooltip reads "Expand files panel"). The sidebar opens into a
285
315
  > **folder tree** (a nested view of your folders and the nodes inside
286
- > them): your five nodes grouped under `.claude/` and `notes/`, each
316
+ > them): your six nodes grouped under `.claude/` and `notes/`, each
287
317
  > row showing its kind and how many links go in and out.
288
318
  >
289
319
  > Tell me when the tree is open.
@@ -292,11 +322,12 @@ Per §Provider detection, on `agent-skills` / Antigravity the fixture has fewer
292
322
 
293
323
  > At the top of that sidebar there's a search box (placeholder
294
324
  > `Search…`). Type `guideline`. Watch both halves at once: the tree
295
- > narrows down to `demo-guideline` and the **Map** drops every node
296
- > except `demo-guideline`. The search matches a node's name, path,
325
+ > narrows down to the two guideline nodes (`demo-guideline` and
326
+ > `demo-guideline2`) and the **Map** drops every node except those
327
+ > two. The search matches a node's name, path,
297
328
  > or description, and filters live as you type, no Enter needed.
298
329
  >
299
- > Now clear the box. All five nodes come back, in both the tree and
330
+ > Now clear the box. All six nodes come back, in both the tree and
300
331
  > the Map. Confirm you saw it filter and then restore.
301
332
 
302
333
  **Beat 3, isolate (tester does this).**
@@ -306,7 +337,8 @@ Per §Provider detection, on `agent-skills` / Antigravity the fixture has fewer
306
337
  > "Isolate this node and its direct links on the map"). Click it.
307
338
  >
308
339
  > The Map collapses to **Demo TODO list** plus only the nodes it
309
- > links to (`demo-command`, `demo-skill`, `demo-guideline`).
340
+ > links to (`demo-command`, `demo-skill`, `demo-guideline`,
341
+ > `demo-guideline2`).
310
342
  > `demo-agent`, which lost its only connector back in the last step,
311
343
  > drops out of view, and the Inspector opens on **Demo TODO list**.
312
344
  > That's how you
@@ -314,7 +346,7 @@ Per §Provider detection, on `agent-skills` / Antigravity the fixture has fewer
314
346
  >
315
347
  > To bring the rest back, look at the toolbar along the bottom of
316
348
  > the Map: there's a **Show all** button (an eye icon). Click it and
317
- > all five nodes return.
349
+ > all six nodes return.
318
350
  >
319
351
  > Did the map isolate and then restore?
320
352
 
@@ -344,14 +376,14 @@ description: |
344
376
  API_TOKEN: example-not-real
345
377
  ```
346
378
 
347
- It lands in the map as a sixth node (`notes/private-credentials`); the watcher sees it like any other `.md`. Do NOT pause to confirm the appearance, it folds into the single vanish confirmation at the end of this step.
379
+ It lands in the map as a seventh node (`notes/private-credentials`); the watcher sees it like any other `.md`. Do NOT pause to confirm the appearance, it folds into the single vanish confirmation at the end of this step.
348
380
 
349
381
  **The tester hides it (single tester-facing message, one confirmation).**
350
382
 
351
383
  Give the tester a mental map of the folder so they know where the file lives, then the glob that hides it, all in ONE message. Use `Bash` (`ls -la`, plus `ls -la notes/` if a deeper view helps) for the real listing and apply the host-dependent rendering rule. Per Inviolable rule #2, the agent does NOT touch `.skillmapignore` with its `Edit` tool, the tester edits it from their own editor:
352
384
 
353
385
  > One last step. Your `private-credentials` note just popped into
354
- > the map as a sixth node, that's the watcher again. Now let's hide
386
+ > the map as a seventh node, that's the watcher again. Now let's hide
355
387
  > it. Here's what your directory looks like right now:
356
388
 
357
389
  ```
@@ -367,6 +399,7 @@ Give the tester a mental map of the folder so they know where the file lives, th
367
399
  └── notes/
368
400
  ├── todo.md
369
401
  ├── demo-guideline.md
402
+ ├── demo-guideline2.md
370
403
  └── private-credentials.md ← what we want to hide
371
404
  ```
372
405
 
@@ -387,8 +420,8 @@ notes/private-*.md
387
420
  >
388
421
  > Watch the browser when you save. The
389
422
  > `notes/private-credentials` node should disappear from the
390
- > **Map** in real time, without restarting anything. Six nodes
391
- > back to five.
423
+ > **Map** in real time, without restarting anything. Seven nodes
424
+ > back to six.
392
425
  >
393
426
  > Did the node vanish?
394
427
 
@@ -1,4 +1,4 @@
1
- # Part 6: Ship the site (the full publish pipeline) (step library, finale, `pipeline` + `golive`)
1
+ # Part 5: Ship the site (the full publish pipeline) (step library, finale, `pipeline` + `golive`)
2
2
 
3
3
  The finale, the climax of the whole campaign. In Part 3 you ran the
4
4
  harness once, the simple way (generate two pages, serve them). Here you
@@ -134,7 +134,7 @@ Tell the tester:
134
134
  > Now take it all in at once. On one side, the real running site you
135
135
  > can click through. On the other, the skill-map graph of the harness
136
136
  > that built it: the handbook, the content editor, the style guide,
137
- > the publish command, the link checker, the MCP tool, all wired
137
+ > the publish command, the link checker, all wired
138
138
  > together. You started in an empty folder with nothing, and you have
139
139
  > ended with a real, running site and a living map of how it all fits.
140
140
  >
@@ -1,6 +1,6 @@
1
1
  # Part 4: Maintain the site (step library, `maintain`)
2
2
 
3
- This is the upkeep part. The harness from Part 2 is wired and clean; real projects drift, links break, drafts pile up, names collide. Here the tester breaks something on purpose and fixes it, meets the analyzer catalogue that catches those problems, finds an orphan nobody links to, clears a reserved-name warning, and learns the `.sm` companion files that carry the tool's bookkeeping. `pace: auto-advance` (walk straight into the next chapter once one is marked done), `preflight: reuse` (it builds on the portfolio harness from Parts 1 and 2, no fresh fixture of its own). Shared conventions (tone, provider detection / substitution, the `> ` rendering rule, the per-step cycle) live in `_core.md`; do not restate them here.
3
+ This is the upkeep part. The harness from Part 2 is wired and clean; real projects drift, links break, drafts pile up, names collide. Here the tester breaks something on purpose and fixes it, meets the analyzer catalogue that catches those problems, finds an orphan nobody links to, clears a reserved-name warning, and learns the `.sm` companion files that carry the tool's bookkeeping. `pace: auto-advance` (walk straight into the next chapter once one is marked done), `preflight: seed` (it builds on the portfolio harness from Parts 1 and 2, reusing the accumulated state when its predecessors are done, no fresh fixture of its own). Shared conventions (tone, provider detection / substitution, the `> ` rendering rule, the per-step cycle) live in `_core.md`; do not restate them here.
4
4
 
5
5
  ## Chapter `broken-ref` - A link breaks (~3 min)
6
6
 
@@ -12,6 +12,21 @@ This chapter is the tester's hands the whole way (renaming and editing their own
12
12
  sm
13
13
  ```
14
14
 
15
+ This part is the first time the live server and one-off `sm` commands run side by side, so give the tester the third-terminal heads-up below (substitute `<cwd>` with the tester's actual cwd, same substitution as every other `<cwd>` mention):
16
+
17
+ > ⚠️ Heads up: from here you run one-off commands (`sm scan`, `sm
18
+ > check`, the `mv` below) **while the `sm` server keeps running**. The
19
+ > server holds its terminal until you Ctrl+C it, so open a **third
20
+ > terminal** and anchor it to this same folder, that is where every
21
+ > one-off command from here on goes:
22
+
23
+ ```bash
24
+ cd <cwd>
25
+ ```
26
+
27
+ > Three terminals now: this chat, the one running the live `sm` server
28
+ > (leave it alone), and this new one for the one-off commands.
29
+
15
30
  > Open the URL `sm` prints, same as before. Now break something on
16
31
  > purpose. Rename your deploy runbook from `docs/DEPLOY.md` to
17
32
  > `docs/DEPLOYMENT.md` (just change the filename, leave the contents
@@ -98,7 +113,7 @@ Wait for confirmation. Mark `analyzers`: done.
98
113
 
99
114
  ## Chapter `orphans` - A page nobody links to (~3 min)
100
115
 
101
- **Context**: a different kind of loose end. A node can be perfectly valid and still be an orphan: nothing in the harness links to it. We create a draft page that no one references, and `sm orphans` finds it. The point is to separate three ideas the tester now has names for: orphan (nothing points at it) vs broken-ref (a link with no target) vs issue (a rule violation).
116
+ **Context**: a different kind of loose end. A node can be perfectly valid and still be an orphan: nothing in the harness links to it. We create a draft page that no one references; it lands on the **Map** as a floating dot, and `sm show` confirms it has no incoming links. The point is to separate three ideas the tester now has names for: orphan (nothing points at it) vs broken-ref (a link with no target) vs issue (a rule violation). Note: orphan-ness is a property of the graph (visible on the Map / inspector), there is no `sm check` analyzer for it. The separate `sm orphans` verb is unrelated, it surfaces history stranded by a rename, not unlinked pages, so do NOT use it here.
102
117
 
103
118
  `Write` `docs/draft.md` (markdown kind), a half-finished page nobody has wired up yet:
104
119
 
@@ -123,19 +138,28 @@ Tell the tester:
123
138
  > I dropped a new note into your project, `docs/draft.md`, a
124
139
  > half-finished page. Look at the **Map**: it shows up as a floating
125
140
  > dot with no arrows touching it. Nothing in your harness links to it,
126
- > which makes it an **orphan**.
141
+ > which makes it an **orphan**. Click the dot, the inspector shows no
142
+ > connections in or out.
143
+ >
144
+ > You can confirm the same thing from the CLI:
127
145
  >
128
146
  > ```bash
129
- > sm orphans
147
+ > sm show docs/draft.md
130
148
  > ```
131
149
  >
132
- > `sm orphans` lists exactly that: nodes nothing points at. It is
133
- > worth keeping three ideas apart, because they are easy to confuse:
150
+ > It prints the node's details, and there is **no "Links in"
151
+ > section**, nothing references it. That absence is exactly what
152
+ > "orphan" means here.
153
+ >
154
+ > It is worth keeping three ideas apart, because they are easy to
155
+ > confuse:
134
156
  >
135
157
  > - **orphan**: a real, valid node that simply has no incoming link
136
- > (your `docs/draft`). Not an error, just unreferenced.
158
+ > (your `docs/draft`). Not an error, just unreferenced, and visible
159
+ > on the Map as a floating dot.
137
160
  > - **broken-ref**: a link whose target file does not exist (the one
138
- > you triggered by renaming the runbook). That is a real issue.
161
+ > you triggered by renaming the runbook). That is a real issue
162
+ > `sm check` reports.
139
163
  > - **issue**: any rule violation `sm check` reports (broken-ref is
140
164
  > one family; name-reserved, self-loop and the rest are others).
141
165
  >
@@ -144,7 +168,14 @@ Tell the tester:
144
168
  > reachable from anywhere. When you link to it later, it stops being
145
169
  > an orphan.
146
170
  >
147
- > Did `sm orphans` list `docs/draft`?
171
+ > (Heads up: there is also an `sm orphans` command, but it is a
172
+ > different tool. It tracks history left stranded when you rename a
173
+ > file that had tracked state, not pages with no links, so it will
174
+ > not list `docs/draft`. Running it here prints "no orphan issues",
175
+ > which is correct.)
176
+ >
177
+ > Did `docs/draft` land as a floating dot, with `sm show` confirming
178
+ > no links in?
148
179
 
149
180
  Wait for confirmation. Mark `orphans`: done.
150
181
 
@@ -167,7 +198,7 @@ args:
167
198
  required: true
168
199
  ---
169
200
 
170
- # /init
201
+ # init
171
202
 
172
203
  Sets up the empty folders a new portfolio needs.
173
204
  ```
@@ -251,36 +282,4 @@ sm sidecar annotate AGENTS.md
251
282
  >
252
283
  > Did the prompt appear, and does `AGENTS.sm` exist now?
253
284
 
254
- Wait for confirmation. You MAY use `Read` on `AGENTS.sm` to verify it landed. Mark `sidecar`: done.
255
-
256
- ## Chapter `versions` - Bump a version, read its history (~3 min)
257
-
258
- **Context**: now that the consent is granted, the day-to-day versioning verbs go through silently. `sm bump` increments a node's frontmatter version and appends a record to its `.sm` companion; `sm history` reads that trail back. We bump the content editor and read its history. Same consent gate as the previous chapter, already satisfied.
259
-
260
- This is a CLI beat, the tester runs everything (substitute `<provider_dir>` in the path per `_core.md`):
261
-
262
- ```bash
263
- sm bump content-editor
264
- sm history content-editor
265
- ```
266
-
267
- > `sm bump <node>` is the everyday versioning verb: it nudges the
268
- > `version` field in the node's frontmatter up by one and appends an
269
- > entry to that node's `.sm` companion file, recording that the bump
270
- > happened. Because you already granted consent in the last chapter,
271
- > it runs without prompting (the `.sm` write is pre-authorized for
272
- > this checkout).
273
- >
274
- > `sm history <node>` reads that trail back: it prints the version
275
- > entries skill-map has recorded for the content editor, so you can
276
- > see how it has changed over time. Right after one bump you will see
277
- > the single entry the bump just wrote.
278
- >
279
- > The two verbs are a pair: `sm bump` writes a version checkpoint into
280
- > the companion file, `sm history` reads the checkpoints back out.
281
- > Your `.md` body never gets cluttered with this, it all lives in the
282
- > `.sm` alongside it.
283
- >
284
- > Does `sm history` show the bump you just made?
285
-
286
- Wait for confirmation. Mark `versions`: done. Last chapter of the part: apply §Closing a part (the close names the part by its title and routes back to the menu).
285
+ Wait for confirmation. You MAY use `Read` on `AGENTS.sm` to verify it landed. Mark `sidecar`: done. Last chapter of the part: apply §Closing a part (the close names the part by its title and routes back to the menu).
@@ -1,10 +1,10 @@
1
- # Part 5: MCP (step library, `mcp-*` ids)
1
+ # Part 8: MCP (step library, `mcp-*` ids)
2
2
 
3
- This is a chapter apart, the one just before the finale. Pace
4
- `auto-advance`, preflight `reuse` (it builds straight on the
5
- connected portfolio harness from the earlier parts, same cwd, same
6
- live `sm` session). One chapter only. Shared conventions live in
7
- `_core.md`.
3
+ This is a chapter apart, a standalone appendix that comes last in the
4
+ book. Pace `auto-advance`, preflight `seed` (`harness-connected`, so it
5
+ fast-forwards onto the connected portfolio harness whether the campaign
6
+ was just run in this directory or you jumped straight here). One chapter
7
+ only. Shared conventions live in `_core.md`.
8
8
 
9
9
  ## Chapter `mcp-node` - The agent reaches for an MCP tool (~3 min)
10
10
 
@@ -1,4 +1,4 @@
1
- # Part 7 (b): Extend skill-map - plugins (step library, `tour-*` ids)
1
+ # Part 6 (b): Extend skill-map - plugins (step library, `tour-*` ids)
2
2
 
3
3
  Guided tour of the **built-in plugins** that ship with `sm`. Three
4
4
  steps: a quick mental model of what plugins are plus a peek at
@@ -13,9 +13,10 @@ which verbs reach which surface.
13
13
  Before announcing the first step, verify the fixture is initialised
14
14
  (the cwd has `.claude/agents/master-agent.md`,
15
15
  `.claude/skills/master-skill/SKILL.md`, AND `.skill-map/` with
16
- `settings.json` and `skill-map.db`). Pre-flight already ran
17
- `sm init --no-scan` and appended the master entries to
18
- `.skillmapignore`. If any of that is missing, surface the
16
+ `settings.json` and `skill-map.db`). The `extend` part's
17
+ `backstage-init` preflight ran `sm init --no-scan` to provision it;
18
+ the universal `.skillmapignore` from pre-flight keeps the tutorial's
19
+ own files out of the scan. If any of that is missing, surface the
19
20
  bootstrap mismatch ("tutorial-state.yml says we are running, but
20
21
  the bootstrap is missing. Re-run the tutorial from an empty dir or
21
22
  restore the files.") and stop.
@@ -5,7 +5,7 @@ starts an actual project: a tiny personal **portfolio website**,
5
5
  fully static, served by a ~15-line Express server, plus the
6
6
  `.claude/` **harness** that maintains it. skill-map maps the harness
7
7
  (the `.md` assets and how they reference each other); the site itself
8
- is plain HTML the harness produces (Part 3 runs it, Part 6 ships it).
8
+ is plain HTML the harness produces (Part 3 runs it, Part 5 ships it).
9
9
  `pace: per-step`, `preflight: portfolio-init`. Shared
10
10
  conventions live in `_core.md`.
11
11
 
@@ -54,8 +54,9 @@ Wait for confirmation. Mark `kickoff`: done.
54
54
 
55
55
  **Context**: the dogfood beat. Real Claude Code projects keep a
56
56
  `CLAUDE.md` that just points at `AGENTS.md` (this very repo does).
57
- That one-line pointer is a real `mentions` link, the tester's first
58
- connector on the real project.
57
+ That one-line pointer is a real `references` link (the `.md` extension
58
+ makes `@AGENTS.md` a file pointer, not a bare mention), the tester's
59
+ first connector on the real project.
59
60
 
60
61
  Tell the tester to create the file themselves (it is their project's
61
62
  file, Inviolable rule #2):
@@ -68,10 +69,13 @@ file, Inviolable rule #2):
68
69
  > ```
69
70
  >
70
71
  > Save it. Watch the map: a new `CLAUDE.md` node appears, with a
71
- > `mentions` connector pointing at `AGENTS.md`. The `@name` token is
72
- > the same mention syntax from the prologue, now doing real work: it
73
- > tells anyone (and skill-map) that `CLAUDE.md` defers to the
74
- > handbook. This is exactly how this tool's own repo is wired.
72
+ > `references` connector pointing at `AGENTS.md`, solid at 1.00.
73
+ > Because `@AGENTS.md` carries the `.md` extension, skill-map reads it
74
+ > as a file pointer (the same `@name.md` reference you met in the
75
+ > prologue, not a bare mention), and since the handbook is right there
76
+ > the link resolves with full confidence. It tells anyone (and
77
+ > skill-map) that `CLAUDE.md` defers to the handbook. This is exactly
78
+ > how this tool's own repo is wired.
75
79
  >
76
80
  > Did the connector light up?
77
81
 
@@ -1,11 +1,12 @@
1
- # Part 3: Run the harness (your site, live) (step library, `generate` + `serve`)
1
+ # Part 3: Run the harness (your site, live) (step library, `generate` + `serve` + `editor-live`)
2
2
 
3
3
  The first payoff: the harness you built and wired in the earlier parts
4
4
  finally does its job and you see a real site running, without waiting
5
5
  for the finale. Pace `auto-advance`, preflight `seed` (`harness-connected`,
6
- so a tester can jump straight here). Two chapters, in order:
7
- `generate` (the agent writes the HTML pages) then `serve` (the tester
8
- runs the site and sees it next to the graph). This is a deliberately
6
+ so a tester can jump straight here). Three chapters: `generate` (the
7
+ agent writes the HTML pages), `serve` (the tester runs the site next to
8
+ the graph), then an optional `editor-live` (the tester lets the real
9
+ `content-editor` agent write a posts page). This is a deliberately
9
10
  simple, working pass: maintenance, MCP and the full publish pipeline
10
11
  come in the parts after it. Shared conventions live in `_core.md`.
11
12
 
@@ -149,7 +150,32 @@ stop the server with Ctrl+C and the edge case for ports applies the
149
150
  same as elsewhere. Remind them they can leave the server running or
150
151
  stop it with Ctrl+C; either way the next parts do not need it.
151
152
 
152
- Mark `serve`: done. Last chapter of the part: apply §Closing a part
153
- (the close names the part by its title and routes back to the menu;
154
- this is a mid-campaign payoff, NOT the campaign finale, so do not
155
- sign the campaign off here).
153
+ Mark `serve`: done. Auto-advance to the optional `editor-live` chapter.
154
+
155
+ ## Chapter `editor-live` - Let the content-editor agent write a posts page for real (optional) (~3 min)
156
+
157
+ **Context**: optional payoff, and the first time the tester runs a harness member for real instead of playing it. In `generate` the tester (as the agent) wrote the HTML by hand; here the actual `content-editor` agent does the job. The tester asks for a new **posts** page, the tutorial invokes the agent, and a real `public/posts.html` lands, proof that the nodes on the map are runnable, not just a diagram. The Layer-1 / Layer-2 split still holds: the new HTML does NOT move the graph (HTML is not `.md`). Fully skippable; run it or skip it, the part closes either way.
158
+
159
+ On `agent-skills` / Antigravity the `content-editor` member is a **skill**, not an `agent`; invoke it as a skill and keep everything else identical.
160
+
161
+ This chapter is OPTIONAL and the tester opts in. Offer it; if they skip, go straight to the part close below.
162
+
163
+ Tell the tester:
164
+
165
+ > Optional last beat, and the fun one: so far you have *played* the
166
+ > `content-editor` yourself. Want to see it run for real? Ask me to add
167
+ > a **posts** page with your agent, for example: "use the
168
+ > content-editor agent to add a posts page". I'll invoke the real
169
+ > `content-editor` in your project; it reads its own rules and the
170
+ > style guide, then writes a new static page into `public/`.
171
+ >
172
+ > Watch two things: the new page lands in `public/` (and shows on the
173
+ > live site when you open it), and the **Map does NOT move**, same
174
+ > Layer-1 / Layer-2 split as before, the agent's HTML output is not
175
+ > part of the harness graph.
176
+ >
177
+ > Or just tell me to skip it and we'll wrap up this part.
178
+
179
+ If the tester opts in: invoke the project's `content-editor` (the `<provider_dir>/agents/content-editor.md` agent, or the skill on `agent-skills`) via the Task tool to write ONE new static page `public/posts.html`, following the agent's own rules and `docs/STYLE.md` (plain HTML, no framework, no client JS, one page per file, a nav link back to Home), holding two or three short sample posts (a heading plus a sentence or two each). Do NOT edit `public/index.html` or any `.md` harness file, and do NOT edit the agent definition. If the subagent is not invocable in the tester's setup, act as the `content-editor` yourself following the same rules so the beat still lands. Then tell the tester to open `http://localhost:3000/posts.html` (refresh and navigate there), confirm the posts page is live and links back home, and confirm the **Map stayed still**.
180
+
181
+ Wait for confirmation (ran it or skipped). Mark `editor-live`: done. Last chapter of the part: apply §Closing a part (the close names the part by its title and routes back to the menu; this is a mid-campaign payoff, NOT the campaign finale, so do not sign the campaign off here).
@@ -1,6 +1,6 @@
1
- # Part 7 (a): Extend skill-map - settings (step library, `settings-*` ids)
1
+ # Part 6 (a): Extend skill-map - settings (step library, `settings-*` ids)
2
2
 
3
- Step bodies for the settings chapters of Part 7 (config layers, the
3
+ Step bodies for the settings chapters of Part 6 (config layers, the
4
4
  `sm config` verbs, the active provider lens), plus the shared step
5
5
  `settings-6-contributions` that the plugin-authoring chapters reuse.
6
6
  The SKILL.md orchestrator dispatches each `settings-*` chapter id
@@ -14,9 +14,10 @@ system and the `sm config` verbs.
14
14
  ## Precondition check
15
15
 
16
16
  Same as the authoring chapters: `.skill-map/` must exist in the
17
- cwd (the orchestrator's backstage-init pre-flight ran `sm init
18
- --no-scan` and appended the tutorial's internal entries to
19
- `.skillmapignore`, so this is the expected state). If
17
+ cwd (the `extend` part's `backstage-init` preflight ran `sm init
18
+ --no-scan` to provision it; the universal `.skillmapignore` from
19
+ pre-flight keeps the tutorial's own files out of the scan, so this
20
+ is the expected state). If
20
21
  `.skill-map/` is missing, surface the bootstrap mismatch and
21
22
  stop, do not try to re-init mid-chapter.
22
23
 
package/dist/cli.js CHANGED
@@ -1,6 +1,6 @@
1
1
  // cli/entry.ts
2
2
 
3
- !function(){try{var e="undefined"!=typeof window?window:"undefined"!=typeof global?global:"undefined"!=typeof globalThis?globalThis:"undefined"!=typeof self?self:{},n=(new e.Error).stack;n&&(e._sentryDebugIds=e._sentryDebugIds||{},e._sentryDebugIds[n]="542dae21-a4ed-50bb-ad3c-62e84e47b380")}catch(e){}}();
3
+ !function(){try{var e="undefined"!=typeof window?window:"undefined"!=typeof global?global:"undefined"!=typeof globalThis?globalThis:"undefined"!=typeof self?self:{},n=(new e.Error).stack;n&&(e._sentryDebugIds=e._sentryDebugIds||{},e._sentryDebugIds[n]="ab2725f4-cc31-54b9-b0c4-0d9778b2143e")}catch(e){}}();
4
4
  import { existsSync as existsSync33 } from "fs";
5
5
  import { Builtins, Cli as Cli2 } from "clipanion";
6
6
 
@@ -246,7 +246,7 @@ function bucketByKind(kind, instance, bag) {
246
246
  // package.json
247
247
  var package_default = {
248
248
  name: "@skill-map/cli",
249
- version: "0.53.3",
249
+ version: "0.53.5",
250
250
  description: "skill-map reference implementation \u2014 kernel + CLI + adapters.",
251
251
  license: "MIT",
252
252
  type: "module",
@@ -30251,4 +30251,4 @@ function resolveBareDefault() {
30251
30251
  process.exit(ExitCode.Error);
30252
30252
  }
30253
30253
  //# sourceMappingURL=cli.js.map
30254
- //# debugId=542dae21-a4ed-50bb-ad3c-62e84e47b380
30254
+ //# debugId=ab2725f4-cc31-54b9-b0c4-0d9778b2143e