@sellable/install 0.1.76 → 0.1.77
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.
|
@@ -20,11 +20,11 @@ Required first steps:
|
|
|
20
20
|
1. Read the three required inputs.
|
|
21
21
|
2. Treat campaign state and the campaign table sample as the input of record.
|
|
22
22
|
Disk files are context/debug aids, not durable state.
|
|
23
|
-
3. Use the embedded Message Review
|
|
24
|
-
long-form `generate-messages` subskill in
|
|
25
|
-
|
|
26
|
-
|
|
27
|
-
|
|
23
|
+
3. Use the embedded Message Review Safety Gate below. Do not load the full
|
|
24
|
+
long-form `generate-messages` subskill in this `create-campaign-v2` path. If
|
|
25
|
+
a needed rule is missing or the draft fails quality gates, return
|
|
26
|
+
`revise-messaging` with the exact failure instead of pulling the long prompt
|
|
27
|
+
into this worker.
|
|
28
28
|
|
|
29
29
|
Owned outputs:
|
|
30
30
|
|
|
@@ -46,8 +46,8 @@ Do not write or modify:
|
|
|
46
46
|
|
|
47
47
|
Process:
|
|
48
48
|
|
|
49
|
-
1. Run the embedded
|
|
50
|
-
lead-review source decision, and `lead-sample.json`.
|
|
49
|
+
1. Run the embedded message-review safety-gate workflow in dry mode from the
|
|
50
|
+
approved brief, lead-review source decision, and `lead-sample.json`.
|
|
51
51
|
2. Use `lead-sample.json` as the only lead sample source. Do not fetch new
|
|
52
52
|
prospects or invent richer row signals.
|
|
53
53
|
3. Build proof inventory, token fill rules, token adherence, angle drafts,
|
|
@@ -69,7 +69,8 @@ Process:
|
|
|
69
69
|
Return a concise final status with:
|
|
70
70
|
|
|
71
71
|
- artifacts written
|
|
72
|
-
- whether embedded
|
|
72
|
+
- whether the embedded message-review safety-gate rules passed or returned
|
|
73
|
+
`revise-messaging`
|
|
73
74
|
- lead sample basis used
|
|
74
75
|
- proposed template and one sample message
|
|
75
76
|
- selected winner summary
|
|
@@ -78,12 +79,12 @@ Return a concise final status with:
|
|
|
78
79
|
Quality bar:
|
|
79
80
|
|
|
80
81
|
- Do not synthesize a lightweight message from general knowledge. The artifact
|
|
81
|
-
must prove the embedded
|
|
82
|
+
must prove the embedded message-review safety-gate workflow ran.
|
|
82
83
|
- Message generation can start before `lead-filter.md`, but message review
|
|
83
84
|
cannot start until the parent verifies the selected basis rows still pass the
|
|
84
85
|
final filter.
|
|
85
86
|
|
|
86
|
-
## Embedded Message Review
|
|
87
|
+
## Embedded Message Review Safety Gate
|
|
87
88
|
|
|
88
89
|
Use this campaign-launch subset to produce a truthful first-send message,
|
|
89
90
|
rendered token examples, and a decision without loading the full long-form
|
package/bin/sellable-install.mjs
CHANGED
|
@@ -843,15 +843,21 @@ updates.
|
|
|
843
843
|
4. For message generation, use the \`post-find-leads-message-scout\` agent when
|
|
844
844
|
available; its prompt carries the campaign-launch message rules. In the
|
|
845
845
|
parent-thread fallback, load
|
|
846
|
-
\`mcp__sellable__get_subskill_asset({ subskillName: "create-campaign-v2", assetPath: "references/message-review-
|
|
846
|
+
\`mcp__sellable__get_subskill_asset({ subskillName: "create-campaign-v2", assetPath: "references/message-review-safety-gate.md" })\`.
|
|
847
|
+
Do not load the full long-form \`generate-messages\` subskill in this
|
|
848
|
+
create-campaign path; if the safety gate cannot safely approve the draft, route
|
|
849
|
+
to revise messaging with the concrete failure.
|
|
847
850
|
Do not synthesize \`message-validation.md\` from the brief, lead review, or
|
|
848
851
|
general knowledge.
|
|
849
|
-
5.
|
|
850
|
-
|
|
851
|
-
the
|
|
852
|
-
|
|
853
|
-
|
|
854
|
-
\`
|
|
852
|
+
5. Create the campaign shell early with the v1 brief so the user can open the
|
|
853
|
+
watch link and see useful setup state immediately. Import only the first
|
|
854
|
+
bounded review batch after the source is attached to the campaign; do not
|
|
855
|
+
queue workflow cells, attach a sequence, or start until
|
|
856
|
+
\`message-validation.md\` proves the message-review safety-gate workflow ran,
|
|
857
|
+
\`message-review.md\` approves the template, rubrics are saved, and the
|
|
858
|
+
approved message set is synced into the campaign brief.
|
|
859
|
+
6. Keep \`selectedLeadListId\` as the source list and \`workflowTableId\` as the
|
|
860
|
+
campaign table. Do not use disk files as the post-mint source of truth.
|
|
855
861
|
7. Do not ask the user to run another command.
|
|
856
862
|
|
|
857
863
|
## Fallback
|
package/package.json
CHANGED
|
@@ -490,14 +490,17 @@ updates.
|
|
|
490
490
|
4. For message generation, use the `post-find-leads-message-scout` agent when
|
|
491
491
|
available; its prompt carries the campaign-launch message rules. In the
|
|
492
492
|
parent-thread fallback, load
|
|
493
|
-
`mcp__sellable__get_subskill_asset({ subskillName: "create-campaign-v2", assetPath: "references/message-review-
|
|
493
|
+
`mcp__sellable__get_subskill_asset({ subskillName: "create-campaign-v2", assetPath: "references/message-review-safety-gate.md" })`.
|
|
494
|
+
Do not load the full long-form `generate-messages` subskill in this
|
|
495
|
+
create-campaign path; if the safety gate cannot safely approve the draft,
|
|
496
|
+
route to revise messaging with the concrete failure.
|
|
494
497
|
Do not synthesize `message-validation.md` from the brief, lead review, or
|
|
495
498
|
general knowledge.
|
|
496
499
|
5. Create the campaign shell early with the v1 brief so the user can open the
|
|
497
500
|
watch link and see useful setup state immediately. Import only the first
|
|
498
501
|
bounded review batch after the source is attached to the campaign; do not
|
|
499
502
|
queue workflow cells, attach a sequence, or start until
|
|
500
|
-
`message-validation.md` proves the
|
|
503
|
+
`message-validation.md` proves the message-review safety-gate workflow ran,
|
|
501
504
|
`message-review.md` approves the template, rubrics are saved, and the
|
|
502
505
|
approved message set is synced into the campaign brief.
|
|
503
506
|
6. Keep `selectedLeadListId` as the source list and `workflowTableId` as the
|