fix: route all scheduled templates to #general (target: channel) + pulse_check chapter limit guard

This commit is contained in:
David Baity
2026-04-28 08:47:55 -04:00
parent 384d43adc4
commit ebc9d1d007
6 changed files with 127 additions and 3 deletions

View File

@@ -0,0 +1,71 @@
name: content_plan
description: "Generate a structured content plan for a Crimson Leaf company or project. Produces a prioritized list of content tasks with assignments and rationale."
debug: true
system: agent_prompt
agent_prompt:
- "= identity.md"
sections:
- agent
- project
- history
- rag
- message
- instructions
steps:
- type: think
hint: |
You are {agent.name}, producing a content plan for Crimson Leaf.
Based on the request and available context, generate a practical content production plan.
Your plan must include:
OBJECTIVE -- One sentence: what this content is for and who it serves.
CONTENT ITEMS -- A numbered list of specific content pieces to produce. For each item:
- Title or working name
- Content type (article, brief, report, script, outline, etc.)
- Target length (approximate word count or format)
- Primary purpose (inform, convert, build authority, entertain, etc.)
- Which agent or team should produce it
- Estimated priority (1=critical, 3=standard, 5=optional)
SEQUENCING -- In what order should items be produced and why?
DEPENDENCIES -- What information, research, or approvals are needed before production starts?
SUCCESS CRITERIA -- How will we know this content plan was effective?
Be specific. Vague content plans produce vague content.
Limit the plan to items that can realistically be executed within two weeks.
- type: package
packet_type: ContentPlanPacket
schema:
objective: "string"
content_items:
- title: "string"
content_type: "string"
target_length: "string"
purpose: "string"
assigned_to: "string"
priority: "integer 1-5"
sequencing_notes: "string"
dependencies: "string"
success_criteria: "string"
insert_children: false
- type: close
rag_update: true
adjudication:
enabled: true
pass_threshold: 65
criteria:
- "Plan includes a clear objective"
- "Each content item has title, type, and assignment"
- "Plan is specific to the request, not generic boilerplate"
- "Success criteria are measurable"

View File

@@ -0,0 +1,42 @@
name: daily_review
description: "Daily operational pulse-check for Crimson Leaf. Reviews company activity, surfaces blockers, and identifies one priority action for the day."
debug: true
system: agent_prompt
agent_prompt:
- "= identity.md"
sections:
- agent
- project
- history
- rag
- message
- instructions
steps:
- type: think
hint: |
You are {agent.name}, conducting the daily Crimson Leaf operational review.
Review the company's current state and produce a concise daily briefing covering:
1. ACTIVITY STATUS -- What tasks completed or are in progress? What is the overall momentum?
2. KEY METRICS -- Any notable numbers: tasks completed this week, companies active, revenue signals.
3. TOP BLOCKER -- What single thing, if resolved, would most improve throughput?
4. PRIORITY ACTION -- One concrete action for today. Be specific: what, who, and why now.
5. FORWARD SIGNAL -- One thing to watch or prepare for in the next 48 hours.
Be direct and brief. This is an executive briefing, not a narrative essay.
Avoid generic observations. Every point should be specific to Crimson Leaf's actual situation.
- type: close
rag_update: true
adjudication:
enabled: true
pass_threshold: 65
criteria:
- "Briefing covers all five required sections"
- "Priority action is concrete and assigned, not vague"
- "Content is specific to Crimson Leaf, not generic business advice"

View File

@@ -70,7 +70,8 @@ steps:
exactly what to do about it. No euphemisms. exactly what to do about it. No euphemisms.
- type: reply - type: reply
target: discussion target: channel
channel_id: "1473516041249493104"
adjudication: adjudication:
enabled: false enabled: false

View File

@@ -66,7 +66,8 @@ steps:
Be bold. Don't hedge. If an opportunity is real, say so with conviction. Be bold. Don't hedge. If an opportunity is real, say so with conviction.
- type: reply - type: reply
target: discussion target: channel
channel_id: "1473516041249493104"
adjudication: adjudication:
enabled: false enabled: false

View File

@@ -56,6 +56,14 @@ steps:
Based on the portfolio report's project_breakdown for crimson_leaf_publishing, Based on the portfolio report's project_breakdown for crimson_leaf_publishing,
which single CLP project has is_idle=true with the most hours_since_activity? which single CLP project has is_idle=true with the most hours_since_activity?
CHAPTER LIMITS -- do NOT dispatch to a book that has reached its chapter target:
- cypress-bend: 12 chapters. If 12 or more polished chapters exist, do NOT dispatch.
- crimson-vows: 10 chapters. If 10 or more polished chapters exist, do NOT dispatch.
- binding-thread: 12 chapters. If 12 or more polished chapters exist, do NOT dispatch.
Check deliverables or portfolio report for polished chapter counts before deciding.
A book at or over its chapter limit should be treated as if is_idle=false.
Valid output values (pick exactly one): cypress-bend, crimson-vows, binding-thread, none Valid output values (pick exactly one): cypress-bend, crimson-vows, binding-thread, none
Rules: output ONLY the slug value. No explanation. No punctuation. No newlines. Rules: output ONLY the slug value. No explanation. No punctuation. No newlines.

View File

@@ -247,7 +247,8 @@ steps:
commit_msg: "log: strategic review cycle {task.id}" commit_msg: "log: strategic review cycle {task.id}"
- type: reply - type: reply
target: discussion target: channel
channel_id: "1473516041249493104"
adjudication: adjudication:
enabled: false enabled: false