@installagent/sdk 0.4.1 → 0.6.0

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.
Files changed (37) hide show
  1. package/dist/agent-mirror.js +22 -10
  2. package/dist/agent-templates.d.ts +10 -12
  3. package/dist/agent-templates.js +34 -6
  4. package/dist/doctor.d.ts +11 -1
  5. package/dist/doctor.js +24 -22
  6. package/dist/index.d.ts +4 -1
  7. package/dist/index.js +4 -1
  8. package/dist/lefthook-hooks.d.ts +36 -0
  9. package/dist/lefthook-hooks.js +64 -0
  10. package/dist/n8n-defaults.d.ts +90 -0
  11. package/dist/n8n-defaults.js +153 -0
  12. package/dist/portal-admin.d.ts +128 -13
  13. package/dist/portal-admin.js +142 -28
  14. package/dist/pr-open.d.ts +45 -0
  15. package/dist/pr-open.js +98 -0
  16. package/dist/sync-repos.d.ts +8 -1
  17. package/dist/sync-repos.js +10 -9
  18. package/dist/triage-enable.js +20 -7
  19. package/package.json +1 -1
  20. package/templates/workspace-root/ONBOARDING.md +9 -7
  21. package/templates/workspace-root/skills/CLAUDE.md +4 -4
  22. package/templates/workspace-root/skills/README.md +2 -2
  23. package/templates/workspace-root/skills/ad-machine/SKILL.md +14 -18
  24. package/templates/workspace-root/skills/backtest-acceptance-gates/SKILL.md +14 -16
  25. package/templates/workspace-root/skills/client-latest/SKILL.md +14 -18
  26. package/templates/workspace-root/skills/email-triage-rollout/SKILL.md +22 -16
  27. package/templates/workspace-root/skills/engagement-spec/SKILL.md +11 -14
  28. package/templates/workspace-root/skills/engagement-spec/references/spec-templates.md +9 -0
  29. package/templates/workspace-root/skills/house-voice/SKILL.md +67 -19
  30. package/templates/workspace-root/skills/ia-cli/SKILL.md +14 -17
  31. package/templates/workspace-root/skills/installagent-brand-kit/SKILL.md +14 -20
  32. package/templates/workspace-root/skills/installagent-brand-kit/scripts/brand.py +2 -2
  33. package/templates/workspace-root/skills/salesforce-api-snapshot/SKILL.md +13 -15
  34. package/templates/workspace-root/skills/slack-voice/SKILL.md +50 -31
  35. package/templates/workspace-root/skills/sso-gated-client-app/SKILL.md +14 -15
  36. package/templates/workspace-root/skills/tech-call-prep/SKILL.md +13 -19
  37. package/templates/workspace-root/skills/vendor-feed-ingestion/SKILL.md +13 -15
@@ -1,23 +1,21 @@
1
1
  ---
2
2
  name: house-voice
3
3
  description: >-
4
- Install Agent's house voice for anything client-facing — proposals, emails,
5
- onboarding/leave-behind guides, web copy, ad copy, Slack replies to a
6
- client. Patterns are pulled verbatim from the installagent.com landing
7
- page, not invented: direct "you/your" address, plain concrete words over
8
- jargon, confident on value but honestly hedged on numbers, the
9
- name-the-objection-then-contrast structure, a one-word-verdict-then-explain
10
- FAQ pattern, and a short list of signature phrases to reuse rather than
11
- paraphrase. Includes a dedicated SLACK section pulled from the account
12
- owner's real client messages — greeting/bullets/ask/enthusiasm skeleton,
13
- his signature phrases, hyphens-not-em-dashes, and emoji as status markers —
14
- which takes precedence over the marketing register when drafting a client
15
- Slack reply. Use this WHENEVER writing or reviewing client-facing sentences,
16
- or someone says "make this sound like us", "in our voice", "match the
17
- website tone", "does this read on-brand", "tighten this copy", "too
18
- jargon-y". Reach for it even without the word "skill". Position-agnostic —
19
- sales, ops, and techs all write client-facing copy. Governs WORDS;
20
- installagent-brand-kit governs visual brand; design-system governs screens.
4
+ Install Agent's house voice for anything a client reads: proposals, emails,
5
+ onboarding guides, web and ad copy, client Slack replies. Patterns come
6
+ verbatim from installagent.com: "you/your" address, plain
7
+ concrete words, confident on value but hedged on numbers,
8
+ name-the-objection-then-contrast, signature phrases to reuse. Sets the
9
+ LENGTH BUDGET for every client message (Slack or email): highlights only,
10
+ ~150 words for an update, max 6 bullets per list, one question, method in a
11
+ thread reply or linked doc. Its SLACK section (the account owner's real
12
+ client messages: greeting/bullets/ask/enthusiasm, hyphens not em dashes,
13
+ emoji as status markers) wins over the marketing register for a client
14
+ Slack reply. Use WHENEVER writing or reviewing client-facing sentences, or
15
+ someone says "in our voice", "does this read on-brand", "tighten this
16
+ copy", "too jargon-y", "too long", "cut the fluff", "highlights only".
17
+ Governs WORDS; installagent-brand-kit governs visual brand; design-system
18
+ governs screens.
21
19
  ---
22
20
 
23
21
  # House voice
@@ -109,6 +107,51 @@ references to an AI system itself.)
109
107
  automation builds." A number or a named institution always beats an
110
108
  adjective like "world-class" used alone.
111
109
 
110
+ ## Highlights only — the length budget
111
+
112
+ Clients read highlights, not the method. A long message is not more
113
+ thorough: the client skims it, misses the ask, and replies late. This applies
114
+ to client Slack and client email alike.
115
+
116
+ **The budget for a client update:**
117
+
118
+ - **~150 words.** Count them before you send. Over budget means detail is
119
+ still in the message.
120
+ - **Max 6 bullets per list**, one fact each. Most updates need two lists or
121
+ fewer.
122
+ - **One question**, on the last line before the close. Two only when both
123
+ block the work.
124
+ - **No item twice.** If a new feature is also a default, say it once.
125
+
126
+ **What stays:** what changed for them (the outcome, not how it works), the
127
+ one number that matters, any default or decision they must confirm, the ask,
128
+ the next step.
129
+
130
+ **What goes** — into a thread reply, the linked doc, or nowhere:
131
+
132
+ - How we calculate it: the rules, the steps, the method
133
+ - Settings variants and "what if" numbers (link them)
134
+ - Why we chose something, unless they asked
135
+ - Internal names, issue ids, caveats on caveats
136
+ - Filler openers: "quick update on where we are and what's left to do",
137
+ "I also want to confirm a few things as we wrap up"
138
+
139
+ The cut test, line by line: **would the client miss it?** If not, delete it.
140
+
141
+ When the client may want the method, send the highlights as the message and
142
+ the method as a thread reply — or keep it ready and send it when they ask.
143
+
144
+ **Before and after** — a real client update, 2026-10-08 (figures removed):
145
+
146
+ | | Before | After |
147
+ |---|---|---|
148
+ | Words | ~400 | ~150 |
149
+ | What's new | 7 items, one repeated in the defaults | 6 items |
150
+ | Defaults to confirm | 3 | 3 |
151
+ | Result | 2 lines + a settings variant | 1 line + a link |
152
+ | Method | 7-step "how we count it" list | thread reply, only if asked |
153
+ | Questions | 2, one buried mid-message | 1, on the last line |
154
+
112
155
  ## Slack, specifically — how the account owner actually writes
113
156
 
114
157
  The landing page is the *marketing* register. Day-to-day client Slack is a
@@ -131,8 +174,10 @@ first and the landing-page pillars second.
131
174
  *"Really enjoyed the call today"*, *"thank you for sending everything over
132
175
  quickly"*
133
176
  3. The substance, as **bullets** (`•`, sub-bullets `◦`) or a numbered list —
134
- never long paragraphs
135
- 4. An explicit, short ask — often literally headed *"Two questions for you:"*
177
+ never long paragraphs, highlights only, inside the
178
+ [length budget](#highlights-only--the-length-budget)
179
+ 4. One explicit, short ask — when there truly are two, he heads them
180
+ *"Two questions for you:"*
136
181
  5. Close on next step + genuine enthusiasm — *"Excited to get going on this"*
137
182
 
138
183
  **Openers actually used:** "Morning [name] -", "Good morning!", "Hi [name]!!",
@@ -209,6 +254,9 @@ news is delivered plainly and immediately, then followed by the plan.
209
254
  ## Before you finish — checklist
210
255
 
211
256
  - [ ] Reads like a person talking to the reader, not a deck abstract.
257
+ - [ ] **If it's a client message (Slack or email)**: words counted, ~150 or
258
+ fewer for an update; max 6 bullets per list; one question; nothing
259
+ said twice; the method moved to a thread reply or a linked doc.
212
260
  - [ ] No banned hype word from the table above.
213
261
  - [ ] At least one claim is backed by a concrete number, name, or example.
214
262
  - [ ] Any bold number/timeline claim is hedged the way the source hedges
@@ -1,23 +1,20 @@
1
1
  ---
2
2
  name: ia-cli
3
3
  description: >-
4
- Route work through the `ia` CLI (team-tools) instead of hand-rolling it —
5
- the decision map from task to command, plus setup/repair via `ia wizard`
6
- and `ia doctor`. Use this WHENEVER a task matches an existing `ia`
7
- operation: answering "what's the status / context on client [slug]?"
8
- (don't grep repos or draft Airtable/Drive/Linear API calls — `ia client`,
9
- `ia client-build`, `ia client-drive`, `ia client-linear` already do it
10
- read-only), starting a new client engagement, building an n8n automation
11
- (check the bank first), running an email-triage gate ("write the soak
12
- report", "sample for gate B", "is the triage agent still recording?" —
13
- `ia triage` reads the classification record and writes the evidence file,
14
- never a hand-built spreadsheet; `ia triage enable` sets a new triage client
15
- up), minting a credential, administering the portal,
16
- searching the team docs, or setting up / fixing a technician's
17
- environment ("set me up", "my env vars are missing", "clone the
18
- workspace", onboarding a new machine). Reach for it before writing any
19
- custom script against GitHub, Airtable, Linear, Drive, n8n, or the
20
- portal — if an `ia` command covers it, use the command.
4
+ Route work through the `ia` CLI (team-tools) instead of hand-rolling it: the
5
+ decision map from task to command, plus setup/repair via `ia wizard` and `ia
6
+ doctor`. Use this WHENEVER a task matches an existing `ia` operation:
7
+ "what's the status / context on client [slug]?" (do not grep repos or
8
+ hand-draft API calls; `ia client`, `ia client-build`, `ia client-drive`, `ia
9
+ client-linear` do it read-only), starting a new client engagement, building
10
+ an n8n automation (check the bank first), running an email-triage gate
11
+ ("write the soak report", "sample for gate B", "is the triage agent still
12
+ recording?"; `ia triage` writes the evidence file from the classification
13
+ record, never a hand-built spreadsheet), minting a credential, administering
14
+ the portal, searching the team docs, or fixing a technician's environment
15
+ ("set me up", "my env vars are missing", "clone the workspace"). Use it
16
+ before writing any custom script against GitHub, Airtable, Linear, Drive,
17
+ n8n, or the portal.
21
18
  ---
22
19
 
23
20
  # ia-cli — use the team CLI, don't reinvent it
@@ -1,26 +1,20 @@
1
1
  ---
2
2
  name: installagent-brand-kit
3
3
  description: >-
4
- The official Install Agent brand kit for deliverables — the canonical palette
5
- (deep red #d11a1a on white / signal red #ff2d2d, white foreground), the official logo
6
- assets, and copy-paste recipes that apply them to Excalidraw diagrams (house
7
- RED theme), client PDFs (SOW, proposal, overview, onboarding, one-pagers), and
8
- **client decks / Google Slides presentations** (scripts/deck.py + a
9
- duplicate-me assets/installagent-deck-template.pptx + scripts/deck_lint.py,
10
- the gate that fails a deck on any off-brand colour, typeface, unresolvable
11
- glyph or out-of-bounds shape and renders every slide so you can actually look
12
- at it). Use this WHENEVER a diagram, document, deck or presentation needs to
13
- look on-brand — triggers include "make it on-brand", "use our brand
14
- color/red", "add our logo", "official logo not a generic one", "branded PDF",
15
- "SOW/proposal PDF", "red excalidraw", "match our theme", "build a deck",
16
- "client presentation", "Google Slides", "slide deck for the call", "deck
17
- template", or any client-facing visual. Reach for it even if the user never
18
- says "skill". It mirrors the single brand source
19
- (team-web/packages/ui/theme.css) and ships scripts/brand.py so generators
20
- never hard-code a hex or re-draw the logo. Pairs with client-workflow-diagram,
21
- engagement-spec, tech-call-prep (the deck's content comes from the call
22
- agenda), and contract-tech-agnostic — this one governs BRAND, those govern
23
- content/layout.
4
+ The official Install Agent brand kit for deliverables: the canonical palette
5
+ (deep red #d11a1a on white, signal red #ff2d2d), the official logo assets,
6
+ and recipes that apply them to Excalidraw diagrams (house RED theme), client
7
+ PDFs (SOW, proposal, overview, onboarding, one-pagers), and client decks /
8
+ Google Slides (scripts/deck.py, a duplicate-me
9
+ assets/installagent-deck-template.pptx, and scripts/deck_lint.py, which
10
+ fails a deck on any off-brand colour, typeface, glyph or out-of-bounds
11
+ shape). Use this WHENEVER a diagram, document, deck or presentation needs to
12
+ look on-brand: "make it on-brand", "use our brand red", "add our logo",
13
+ "branded PDF", "SOW/proposal PDF", "red excalidraw", "build a deck", "client
14
+ presentation", "Google Slides", or any client-facing visual. Mirrors
15
+ team-web/packages/ui/theme.css; ships scripts/brand.py so generators never
16
+ hard-code a hex. This one governs BRAND; client-workflow-diagram,
17
+ engagement-spec and contract-tech-agnostic govern content/layout.
24
18
  ---
25
19
 
26
20
  # Install Agent brand kit — branded diagrams & PDFs
@@ -128,8 +128,8 @@ def pdf_css(size="Letter", accent=BRAND) -> str:
128
128
  return f"""
129
129
  @page {{ size: {size}; margin: 16mm 17mm 14mm 17mm;
130
130
  @top-left {{ content: element(rhead); }}
131
- @bottom-left {{ content:"Install Agent · Confidential"; font-size:8pt; color:{FAINT}; }}
132
- @bottom-right {{ content:"Page " counter(page) " of " counter(pages); font-size:8pt; color:{FAINT}; }} }}
131
+ @bottom-left {{ content:"Install Agent · Confidential"; font-family:{FONT_BODY}; font-size:8pt; color:{FAINT}; }}
132
+ @bottom-right {{ content:"Page " counter(page) " of " counter(pages); font-family:{FONT_BODY}; font-size:8pt; color:{FAINT}; }} }}
133
133
  @page:first {{ @top-left {{ content: none }} }}
134
134
  * {{ box-sizing:border-box }}
135
135
  body {{ margin:0; color:{BODY}; font-family:{FONT_BODY}; font-size:10.3pt; line-height:1.5; }}
@@ -2,21 +2,19 @@
2
2
  name: salesforce-api-snapshot
3
3
  description: >-
4
4
  Read data out of a client's Salesforce org (including managed-package orgs
5
- like Lumary, Vlocity, nCino) and get it into something we build on — by
6
- discovering the real object model with Tooling API SOQL instead of clicking
7
- through Object Manager, then caching a scheduled read as a JSON snapshot in
8
- GCS/S3 rather than querying Salesforce live from the app. Covers the four
9
- things that waste a day each: the managed-package namespace trap (Object
10
- Manager's search matches Labels, not API names, so it undercounts by 40x),
11
- telling a real lookup from an external-ID text field (which is how you find
12
- out the data you need isn't in Salesforce at all), the three-layer seam that
13
- lets you build the entire pipeline against invented fixtures before you ever
14
- hit a production org with no sandbox, and the least-privilege bucket wiring
15
- between the n8n writer and the app reader. Use this WHENEVER an engagement
16
- reads Salesforce (or any vendor CRM/EHR API) — "connect to their Salesforce",
17
- "pull data from Lumary", "we need their CRM data in the dashboard", "what
18
- objects do they have", "build the ingest", "cache the API response" — and
19
- whenever a build has to run against a production org with no sandbox.
5
+ like Lumary, Vlocity, nCino) into something we build on: discover the real
6
+ object model with Tooling API SOQL, not Object Manager clicks, then cache a
7
+ scheduled read as a JSON snapshot in GCS/S3 instead of querying Salesforce
8
+ live from the app. Covers the four things that waste a day each: the
9
+ managed-package namespace trap (Object Manager search matches Labels, not
10
+ API names, so it undercounts by 40x), telling a real lookup from an
11
+ external-ID text field, the three-layer seam that lets you build the whole
12
+ pipeline on invented fixtures before touching a production org with no
13
+ sandbox, and least-privilege bucket wiring between the n8n writer and the
14
+ app reader. Use this WHENEVER an engagement reads Salesforce (or any vendor
15
+ CRM/EHR API): "connect to their Salesforce", "pull data from Lumary", "we
16
+ need their CRM data in the dashboard", "build the ingest", or "cache the API
17
+ response".
20
18
  ---
21
19
 
22
20
  # Salesforce API → cached snapshot
@@ -1,29 +1,30 @@
1
1
  ---
2
2
  name: slack-voice
3
3
  description: >-
4
- Write Slack messages the Install Agent team will actually read: terse, factual,
5
- bulleted, hard numbers, no prose paragraphs. Use this WHENEVER you are about to
6
- send or draft a Slack message, DM, or channel post — a status update, a PR
7
- notification, a question for a teammate, a client-channel reply, a stand-up
8
- note, "tell Angello X", "post an update in #team-techs", "dm Koz", "let the
9
- team know". Reach for it even if the user never says "skill": any time the
10
- output is going into Slack, this is the skill. It exists because a ~400-word
11
- DM on 2026-08-11 got the reply "Bro, can we please use english" and "I will not
12
- read all this everytime". Length is not thoroughness in Slack — it is a
13
- delivery failure. INTERNAL Slack only — team channels and teammate DMs. For a
14
- Slack message a CLIENT will read, use house-voice instead; its Slack section
15
- is the client register (greeting, bullets, explicit ask, warmth) and it wins
16
- there. Does NOT apply to repo docs, evidence/ files, commit bodies, PR
17
- descriptions, or Linear issues; those stay detailed.
4
+ Write internal messages the Install Agent team will actually read: terse,
5
+ factual, bulleted, hard numbers, highlights only, ~100 words max for an
6
+ update. Use WHENEVER you draft a message a teammate reads: a Slack DM or
7
+ channel post, an email, a Google Chat message, a status update, a PR
8
+ notification, a stand-up note, "tell Angello X", "post in #team-techs",
9
+ "dm Koz", "email the team", "let the team know", "this is too long", "cut
10
+ the fluff". Reach for it even if the user never says "skill". It exists
11
+ because a ~400-word DM on 2026-08-11 got the reply "Bro, can we please use
12
+ english". Length is not thoroughness in a message — it is a delivery
13
+ failure. INTERNAL only. For a message a CLIENT reads, use house-voice: its
14
+ length budget and Slack section win there. Does NOT apply to repo docs,
15
+ evidence/ files, commit bodies, PR descriptions, or Linear issues; those
16
+ stay detailed.
18
17
  ---
19
18
 
20
- > **Which skill applies.** Internal teammate → this file. Client → `house-voice`
21
- > § *Slack, specifically*. The registers are different on purpose: a client gets
22
- > a greeting and an explicit ask; Angello gets a link and a number.
19
+ > **Which skill applies.** Internal teammate (Slack, email, chat) → this file.
20
+ > Client → `house-voice` § *Highlights only* and § *Slack, specifically*. The
21
+ > registers are different on purpose: a client gets a greeting and an explicit
22
+ > ask; Angello gets a link and a number.
23
23
 
24
24
  # Skill: slack-voice
25
25
 
26
- **Use when:** anything you write is going into Slack.
26
+ **Use when:** anything you write is a message a teammate reads — Slack, email,
27
+ or chat. The name says Slack because that is where most of it goes.
27
28
 
28
29
  ## The bar
29
30
 
@@ -41,26 +42,32 @@ add only what carries information.
41
42
  ## Rules
42
43
 
43
44
  1. **Lead with the link or the number.** No preamble. The first line is the
44
- payload.
45
- 2. **30 words per sentence, maximum.** Count them. Most should be under 15.
46
- 3. **Bullets, not paragraphs.** One fact per bullet. Cut any clause that
47
- explains why.
48
- 4. **Every claim carries a figure.** `18 of 19 checks pass`, `0 client cells
45
+ payload. Delete openers like "Hey, quick update on where we are", "Just
46
+ wanted to flag", "I also want to confirm a few things".
47
+ 2. **~100 words per message, maximum, for an update.** Count them. Over 100
48
+ means the method is still in the message — move it to the thread or the
49
+ file you link. A one-line answer needs no budget.
50
+ 3. **Highlights only.** The reader gets the outcome and the ask, not how you
51
+ got there. Line by line: would the reader miss it? If not, delete it.
52
+ 4. **30 words per sentence, maximum.** Count them. Most should be under 15.
53
+ 5. **Bullets, not paragraphs.** One fact per bullet, max 6 per list, nothing
54
+ said twice. Cut any clause that explains why.
55
+ 6. **Every claim carries a figure.** `18 of 19 checks pass`, `0 client cells
49
56
  touched`, `6 days left`. A number can be checked. A description cannot.
50
- 5. **No adjectives, no adverbs, no fillers.** Delete: *solid, clean, robust,
57
+ 7. **No adjectives, no adverbs, no fillers.** Delete: *solid, clean, robust,
51
58
  healthy, significant, comprehensive, nicely, properly, actually, basically,
52
59
  just, really, quite, worth noting, in good shape.* If a number cannot replace
53
60
  the word, delete the sentence.
54
- 6. **No jargon.** Write what a reader says out loud, not what the code calls it.
61
+ 8. **No jargon.** Write what a reader says out loud, not what the code calls it.
55
62
  "Two approvals at once both write" — not "`allocate` returns `reused:true`
56
63
  before `assertStillOurs()`". Keep an identifier only when the reader needs it
57
64
  to search: `PR #8`, `INS-202`, a file name, a branch.
58
- 7. **No reasoning, no hedging, no advocacy.** Cut "I think", "arguably", "where
65
+ 9. **No reasoning, no hedging, no advocacy.** Cut "I think", "arguably", "where
59
66
  I'd still push". State the conclusion. Link the argument.
60
- 8. **No section headers in a short message.** Bold labels and `###` are what
61
- make a message read as a wall. One bold first line is the maximum.
62
- 9. **Ask one question, not three.** Three answers means three messages.
63
- 10. **Draft before sending to a channel.** Show the user first when the audience
67
+ 10. **No section headers in a short message.** Bold labels and `###` are what
68
+ make a message read as a wall. One bold first line is the maximum.
69
+ 11. **Ask one question, not three.** Three answers means three messages.
70
+ 12. **Draft before sending to a channel.** Show the user first when the audience
64
71
  is the team or a client. A DM to the user themselves can go direct.
65
72
 
66
73
  ## Sending something over
@@ -73,6 +80,9 @@ Put the thing in the message. Put the context in the thread.
73
80
  This keeps the channel scannable. Someone who only needs the link stops at the
74
81
  first message. Someone who needs the detail opens the thread.
75
82
 
83
+ **Email to a teammate:** same budget. The subject line is the point. The body
84
+ is the highlights. The detail is a link or an attachment, not more paragraphs.
85
+
76
86
  ## Shape that works
77
87
 
78
88
  ```
@@ -128,13 +138,22 @@ nobody re-asked, and three paragraphs of reasoning that belonged in the
128
138
  `evidence/` file the message could have linked to.
129
139
 
130
140
  The sent version still carried jargon — `410 on reuse`, `schedule_not_warmed`,
131
- `allocate` — which rule 6 now removes. Compare the two bullets:
141
+ `allocate` — which rule 8 now removes. Compare the two bullets:
132
142
 
133
143
  | Sent | Better |
134
144
  |---|---|
135
145
  | Approval links are single-use: commit once, 410 on reuse. | An approval link works once. The second use is refused. |
136
146
  | Unwired project holds as `schedule_not_warmed`. | A project we have not set up yet stops, and says why. |
137
147
 
148
+ ## Before you send
149
+
150
+ - [ ] First line is the link, the number, or the point — no opener.
151
+ - [ ] Words counted: ~100 or fewer for an update.
152
+ - [ ] Max 6 bullets per list. Nothing said twice.
153
+ - [ ] One question.
154
+ - [ ] Method, reasoning, and "what if" numbers moved to the thread or a linked
155
+ file.
156
+
138
157
  ## Channel rules that still apply
139
158
 
140
159
  - **`#client-*` channels have the client in them.** DM for internal review.
@@ -1,21 +1,20 @@
1
1
  ---
2
2
  name: sso-gated-client-app
3
3
  description: >-
4
- Stand up an internal web app for a client — a dashboard, map, worklist, or
5
- review screen — that only that client's staff can open, using Google Workspace
6
- SSO on a Next.js app deployed to Vercel. Covers the two-layer auth gate
7
- (middleware/proxy for pages + an independent session check inside every API
8
- route), the hosted-domain check that actually restricts sign-in, the
9
- narrower email allowlist for extra-sensitive data, and the four failure modes
10
- that all look like "it works" while the app is wide open: a missing
11
- AUTH_SECRET failing OPEN rather than closed, a proxy matcher that swallows
12
- API routes and hands fetch() an HTML login page, a dev bypass wired to a
13
- custom env var instead of NODE_ENV, and a deployed app whose Vercel env vars
14
- were never set. Use this WHENEVER an engagement produces a screen a client
15
- logs into — "build them a dashboard", "host the map", "an internal tool for
16
- their team", "put it behind their Google login", "only their staff should see
17
- this" — and whenever verifying that an already-deployed client app is
18
- actually gated.
4
+ Stand up an internal web app for a client (dashboard, map, worklist, or
5
+ review screen) that only that client's staff can open, using Google
6
+ Workspace SSO on a Next.js app deployed to Vercel. Covers the two-layer auth
7
+ gate (middleware/proxy for pages + an independent session check inside every
8
+ API route), the hosted-domain check that actually restricts sign-in, the
9
+ narrower email allowlist for extra-sensitive data, and the four failure
10
+ modes that look like "it works" while the app is wide open: a missing
11
+ AUTH_SECRET failing OPEN, a proxy matcher that swallows API routes and hands
12
+ fetch() an HTML login page, a dev bypass wired to a custom env var instead
13
+ of NODE_ENV, and Vercel env vars never set on the deployed app. Use this
14
+ WHENEVER an engagement produces a screen a client logs into: "build them a
15
+ dashboard", "host the map", "an internal tool for their team", "put it
16
+ behind their Google login", "only their staff should see this", or you must
17
+ verify a deployed client app is gated.
19
18
  ---
20
19
 
21
20
  # SSO-gated internal client app
@@ -1,25 +1,19 @@
1
1
  ---
2
2
  name: tech-call-prep
3
3
  description: >-
4
- Verify a technician is actually prepared for an upcoming client call and
5
- generate the per-client call agenda they run it from — the full
6
- run-of-show (read-back facts, walkthrough plan, tailored questions,
7
- access table, expectations, close checklist), not just a question list.
8
- Reads the registry (team-clients/clients/[slug]/) and the build repo
9
- (clients/client-[slug]/ docs 01–04, onboarding.html), checks the
10
- come-prepared checklist (repo scaffolded, engagement docs drafted,
11
- proposals loaded), pulls in the auto-drafted agenda Doc at the root of
12
- the client's Drive folder when one exists (the portal Lead Close Kickoff
13
- creates it when sales closes the lead), and writes docs/05-call-agenda.md
14
- (superseding the old 05-call-sheet.md) plus a Google Doc copy at the
15
- root of the client's Drive folder. Use this WHENEVER a discovery/scoping call is coming
16
- up, or someone says "prep me for the call with [client]", "generate the
17
- call agenda / call sheet", "am I ready for the [client] call?", or you're
18
- working in a client-[slug]/ repo before a call. Reach for it even if the
19
- user never says "skill" — turning the registry profile + call history +
20
- engagement docs into a concrete plan for the next conversation. The
21
- static call script lives on the team hub (Techs → Discovery call); this
22
- skill produces the per-client half.
4
+ Verify a technician is prepared for an upcoming client call and generate the
5
+ per-client call agenda they run it from: the full run-of-show (read-back
6
+ facts, walkthrough plan, tailored questions, access table, expectations,
7
+ close checklist). Reads the registry (team-clients/clients/[slug]/) and the
8
+ build repo (clients/client-[slug]/ docs 01–04, onboarding.html), checks the
9
+ come-prepared checklist (repo scaffolded, engagement docs drafted, proposals
10
+ loaded), pulls in the auto-drafted agenda Doc from the client's Drive folder
11
+ when one exists, and writes docs/05-call-agenda.md plus a Google Doc copy in
12
+ that folder. Use this WHENEVER a discovery/scoping call is coming up, or
13
+ someone says "prep me for the call with [client]", "generate the call agenda
14
+ / call sheet", "am I ready for the [client] call?", or you are in a
15
+ client-[slug]/ repo before a call. The static call script lives on the team
16
+ hub (Techs → Discovery call); this skill produces the per-client half.
23
17
  ---
24
18
 
25
19
  # Tech call prep — the per-client call agenda
@@ -1,22 +1,20 @@
1
1
  ---
2
2
  name: vendor-feed-ingestion
3
3
  description: >-
4
- Get a client's vendor data arriving automatically every day — via our shared
5
- SFTP drop (SFTPGo → S3) or the data@ ingest inbox — and make the pipeline
4
+ Get a client's vendor data arriving automatically every day, via our shared
5
+ SFTP drop (SFTPGo → S3) or the data@ ingest inbox, and make the pipeline
6
6
  refuse to publish garbage. Covers the source-preference ladder (API →
7
- scheduled vendor push → RPA last resort, with manual export as bootstrap
8
- only), the delivery contract you hand the person building the push
9
- (atomic write, one stable path, cadence, required columns, join key), the
10
- format traps that eat a morning each (UTF-8 BOM, .xls named .csv, column-name
11
- variants, invented status vocabularies, uploads landing in a nested path),
12
- the difference between a snapshot export and a change-log export, and — the
13
- part nobody builds until it's too late — two-tier feed guards calibrated
14
- against real versioned deliveries, tuned so they never fire on a normal
15
- morning and never blow away yesterday's good output. Use this WHENEVER an
16
- engagement ingests a recurring file or export — "set up the daily feed",
17
- "they'll send us a CSV", "pull their export", "the data looks wrong today",
18
- "the workflow ran but the numbers are off", "onboard a new client feed", or
19
- provisioning an SFTP user.
7
+ scheduled vendor push → RPA last resort, manual export as bootstrap only),
8
+ the delivery contract you hand the person building the push (atomic write,
9
+ one stable path, cadence, required columns, join key), the format traps that
10
+ eat a morning each (UTF-8 BOM, .xls named .csv, column-name variants,
11
+ invented status vocabularies, nested upload paths), snapshot vs change-log
12
+ exports, and two-tier feed guards calibrated against real versioned
13
+ deliveries that never fire on a normal morning and never blow away
14
+ yesterday's good output. Use this WHENEVER an engagement ingests a recurring
15
+ file or export: "set up the daily feed", "they'll send us a CSV", "pull
16
+ their export", "the data looks wrong today", "the workflow ran but the
17
+ numbers are off", or provisioning an SFTP user.
20
18
  ---
21
19
 
22
20
  # Vendor feed ingestion