Delivery Skills Platform Plan
Date: 2026-03-01 Author: Delivery Platform Team Source: “Forging the Future” role definitions, ops sync transcripts, Brile’s EP process work Purpose: Map every recurring CSO, EP, and SL task to a Cursor skill, MCP, or platform integration that the delivery platform team should build so leadership overlays start with a fully equipped toolbox.
1. Current State
What exists today
| Asset | Type | Serves |
|---|---|---|
meeting-prep | Cursor skill | EP, CSO — pull recent vault context before a call |
humanizer | Cursor skill | All — clean AI writing from outbound text |
linear-audit (Linear board audit) | Cursor command + rule | Create/update Linear tickets from transcripts (full mode) or summary-only (standard) |
enrich-ops-ticket | Cursor command + rule | Ops — add knowledge/Notion context to a Linear issue |
| Linear MCP | MCP server | EP, SL — read/write Linear issues |
| Notion MCP | MCP server | EP, CSO — read Notion pages |
| Google Workspace MCP | MCP server | All — calendar, docs, email |
| Granola MCP | MCP server | All — meeting notes |
| Figma MCP | MCP server | SL (design) |
| Operating MCP | MCP server | Ops — Operating.app |
| Hex MCP | MCP server | Data — analytics |
| GTM agent workers (14) | Vault agents | GTM — campaigns, outreach, analysis |
| Platform (The Forge) | Next.js app | Transcripts, deals, client hub, AI chat |
What Brile has started (EP-focused)
- EP audit Cursor command (shipped)
- Templated messages for weekly EP tasks (in progress)
- Cursor onboarding guide for new hires (in progress)
- Time estimates: ~30 min/client/day target for EP work
Gaps
- CSO and SL have zero dedicated skills today.
- No skill automates the weekly client presentation or health report.
- No skill handles standup prep, standup recap, or daily client pulse.
- No Instagant/Gantt integration exists in Cursor.
- Slack messages are not in the vault (only transcripts).
- No automated KPI tracking (communication consistency, defect density, etc.).
- No asset codification workflow.
2. Role Task Matrix
Derived from “Forging the Future” sections 04-06 and the Day-in-the-Life drafts.
CSO — Customer Success Owner
| Cadence | Task | Time Est. | Automatable? |
|---|---|---|---|
| Daily | Client pulse (personalized Slack/email to client) | 15 min/client | High — draft from yesterday’s standup + transcript |
| Daily | Stand-up advocacy (report client sentiment, expansion signals) | 5 min/client | Medium — prep notes from transcript + Linear |
| Daily | Internal health board update | 10 min | High — auto-generate from Linear + transcript |
| Weekly | Weekly client presentation (value showcase) | 90 min | High — auto-generate deck from week’s work |
| Weekly | Team delivery sync (co-lead with EP) | 30 min | Medium — prep agenda + talking points |
| Weekly | Client health report (KPIs, milestones, ticket status) | 30 min | High — pull from Linear + Notion + transcripts |
| Weekly | Expansion value pipeline review | 15 min | Medium — surface signals from transcripts |
| Monthly | Value realization audit (output vs. business goals) | 60 min | Medium — compare deliverables to SOW goals |
| Monthly | Internal retrospective with SL | 30 min | Low — human conversation, but prep can be automated |
| Monthly | AI efficiency audit (own CSO tasks) | 20 min | Medium — usage analytics |
| Monthly | NPS / sentiment review | 20 min | High — aggregate from transcripts + health board |
| Quarterly | Executive Business Review (EBR) | 2-3 hr | Medium — deck generation, data pull |
| Quarterly | Expansion roadmapping | 1-2 hr | Medium — pattern analysis from transcripts |
| Lifecycle | Kickoff (scope, Gantt, resource plans with EP) | 4-6 hr | Low — human-led, but templates help |
| Lifecycle | Close-out (case study, final report, renewal setup) | 3-4 hr | Medium — case study draft, final impact report |
EP — Engagement Planner
| Cadence | Task | Time Est. | Automatable? |
|---|---|---|---|
| Daily | Ticket audit (every item has owner + priority) | 15 min/client | High — Linear board scan + flag gaps |
| Daily | Priority sync with SL | 10 min | Medium — prep blockers list from Linear |
| Daily | Internal health board update | 10 min | High — auto from Linear status |
| Weekly | Internal delivery sync (co-lead with CSO) | 30 min | Medium — prep agenda from week’s data |
| Weekly | Roadmap calibration (update Gantt/milestones) | 30 min | Medium — need Gantt integration |
| Weekly | Resource assessment (flag capacity bottlenecks) | 20 min | Medium — compare velocity to remaining scope |
| Weekly | Client health report update | 20 min | High — pull from Linear + Notion |
| Monthly | Operational retrospective | 45 min | Medium — prep from transcript themes |
| Monthly | Capacity forecasting | 30 min | Medium — need resource/allocation data |
| Quarterly | Blueprint phase (project plan, Gantt, resource strategy) | 4-6 hr | Low — templates help |
| Lifecycle | Close-out (post-mortem, archive artifacts, template creation) | 2-3 hr | Medium — auto-archive + template extraction |
SL — Service Leader
| Cadence | Task | Time Est. | Automatable? |
|---|---|---|---|
| Daily | Host standup (facilitate 4-min heartbeat per client) | 15 min/client | Medium — prep talking points, post recap |
| Daily | Functional oversight (review complex work items) | 30 min | Medium — pull PRs + Linear tickets for review |
| Daily | Code quality review (previous day’s output) | 30 min | Medium — PR summary + defect flagging |
| Daily | Tomorrow’s solutioning (pre-solve hard tickets) | 30 min | Low — human judgment, but context prep helps |
| Weekly | Technical quality audit (project artifacts review) | 45 min | Medium — automated checklist from Linear/GitHub |
| Weekly | Internal delivery sync with EP | 30 min | Medium — prep blockers + technical debt summary |
| Weekly | Solutioning leadership (direct technical strategy) | 30 min | Low — human-led |
| Weekly | Asset hunt (identify reusable patterns from week’s code) | 20 min | Medium — scan PRs for template candidates |
| Monthly | Asset codification (convert patterns to templates/agents) | 2-3 hr | Medium — scaffold from identified patterns |
| Monthly | Junior velocity audit (+15%/quarter tracking) | 30 min | High — pull Linear velocity metrics |
| Monthly | AI efficiency audit (technical AI ratio) | 20 min | Medium — usage analytics |
| Quarterly | Talent calibration (recruitment support, performance) | 1-2 hr | Low — human judgment |
| Lifecycle | Technical discovery (architecture, SME identification) | 4-6 hr | Low — templates help |
| Lifecycle | Quality gate (final technical sign-off) | 1-2 hr | Medium — automated checklist |
3. Proposed Skills to Build
Each skill below is a .cursor/skills/{name}/SKILL.md file that the delivery platform team develops and ships. Skills are ordered by impact and grouped by role.
Tier 1 — High Impact, Ship First (8 skills)
These address the most frequent tasks (daily/weekly) with the highest automation potential.
1. client-pulse (CSO)
What it does: Drafts a personalized daily client touch message (Slack or email) using yesterday’s standup transcript, recent Linear activity, and any pending action items. Inputs: Client name Outputs: Draft message ready to send MCPs needed: Linear, Google Workspace (email), Granola (if standup was recorded) Data sources: Vault transcripts, Linear issues, previous pulse messages
2. standup-prep (CSO + SL + EP)
What it does: Generates each role’s standup talking points in the Heartbeat Protocol format (SL: technical story; CSO: partnership pulse; EP: guardrail check; SL: plan of attack). Inputs: Client name, role Outputs: Structured 60-second talking points per role MCPs needed: Linear Data sources: Linear board, vault transcripts (last 24h), GitHub PRs (for SL)
3. standup-recap (All)
What it does: After a standup, parses the transcript and produces a structured recap in the Heartbeat Protocol format. Posts to the client channel or saves to vault. Inputs: Transcript (auto or manual) Outputs: Formatted recap with action items, blockers, sentiment MCPs needed: Linear (to create tickets from action items), Notion (to update health board) Data sources: Granola/vault transcript
4. weekly-client-presentation (CSO)
What it does: Generates a weekly value presentation deck from the week’s Linear tickets, merged PRs, standup recaps, and client feedback. Follows Brainforge deck template. Inputs: Client name, week range Outputs: Slide deck (Google Slides or Markdown) with “hero wins,” progress metrics, and next steps MCPs needed: Linear, Google Workspace (Slides), GitHub (PR summaries) Data sources: Linear completed tickets, vault transcripts, GitHub PRs
5. client-health-report (CSO + EP)
What it does: Compiles the formal weekly/monthly health report covering KPIs, milestone status, ticket summary, sentiment score, and risk flags. Inputs: Client name, period (week/month) Outputs: Structured health report (Markdown or Notion page) MCPs needed: Linear, Notion Data sources: Linear metrics, vault transcripts, Notion project pages, previous health reports
6. ticket-quality-audit (EP — extends existing linear-audit)
What it does: Scans the Linear board and flags: tickets without owners, tickets without priorities, stale tickets, tickets missing descriptions, started tickets without assignees, cycle-push candidates. Inputs: Client name or Linear team Outputs: Audit report with fix suggestions; optionally auto-fix (add owners, priorities) MCPs needed: Linear Data sources: Linear board state
7. code-quality-digest (SL)
What it does: Summarizes the previous day’s (or week’s) PRs: what was merged, review comments, defect density, test coverage changes, and any architectural concerns. Inputs: Client name or GitHub repo Outputs: Structured digest with defect density metric, PR summaries, review highlights MCPs needed: GitHub (new — see MCPs section below) Data sources: GitHub PRs, review comments, CI status
8. blocker-radar (EP + SL)
What it does: Scans Linear for blocked/stalled tickets, cross-references with standup transcripts to identify recurring blockers, and produces a prioritized unblock list with suggested actions. Inputs: Client name Outputs: Prioritized blocker list with context and suggested resolution paths MCPs needed: Linear Data sources: Linear issues (blocked status, stale), vault transcripts
Tier 2 — Medium Impact, Ship Second (7 skills)
9. expansion-signal-scanner (CSO)
What it does: Scans recent transcripts and client communications for expansion signals (new feature requests, adjacent problems mentioned, budget hints, new stakeholder introductions) and surfaces them as opportunities. Inputs: Client name, time range Outputs: List of expansion signals with context quotes and suggested next steps
10. value-realization-audit (CSO)
What it does: Compares delivered work (completed Linear tickets, shipped features) against the original SOW objectives and produces a gap analysis with ROI narrative. Inputs: Client name Outputs: SOW-to-delivery mapping, completion %, value narrative Data sources: SOW (vault), Linear completed work, vault transcripts
11. retrospective-prep (CSO + SL)
What it does: Prepares a structured retrospective agenda by analyzing the month’s transcripts, health reports, and ticket data for friction themes, wins, and improvement areas. Inputs: Client name, period Outputs: Retro agenda with categorized themes (What went well / What didn’t / Action items)
12. ebr-deck-generator (CSO)
What it does: Generates the quarterly Executive Business Review deck: ROI summary, milestone timeline, NPS trend, expansion roadmap, next-quarter plan. Inputs: Client name, quarter Outputs: EBR slide deck MCPs needed: Linear, Google Workspace (Slides), Notion
13. resource-capacity-forecast (EP)
What it does: Projects resource needs based on current velocity, remaining scope, and upcoming milestones. Flags over/under-allocation. Inputs: Client name or project Outputs: Capacity report with risk flags Data sources: Linear velocity, milestone dates, team allocation data
14. asset-codification-scout (SL)
What it does: Scans the week’s/month’s code for reusable patterns, prompt chains, or configurations that could become Brainforge templates or agents. Produces a candidate list with extraction instructions. Inputs: GitHub repo(s) Outputs: Asset candidate list with code references and extraction plan MCPs needed: GitHub
15. junior-velocity-tracker (SL)
What it does: Pulls Linear velocity metrics per team member over time, calculates the +15%/quarter target, and flags team members who may need additional mentorship. Inputs: Linear team, time range Outputs: Velocity report per person with trend analysis
4. MCPs and Integrations to Build
New MCPs needed
| MCP | Purpose | Serves | Priority |
|---|---|---|---|
| GitHub MCP | Read PRs, reviews, CI status, commit history. Needed for code-quality-digest, asset-codification-scout, weekly-client-presentation. | SL, EP | P0 |
| Instagant / Gantt MCP | Read/update Gantt charts and milestones. Brile flagged Instagant as a key EP tool that is currently disconnected from Cursor. | EP | P1 |
| Slack MCP | Read/search Slack messages. Transcripts come into vault, but Slack messages do not. CSO daily pulses happen in Slack; EP scope discussions happen in Slack. Having Slack context in Cursor enables better prep and search. | CSO, EP, SL | P1 |
| HubSpot MCP | Read deal data, log activities, pull pipeline info. Currently only an API route in the platform; a Cursor MCP would let CSOs check deal/renewal status directly. | CSO | P2 |
| Clockify MCP | Read time entries. Useful for resource-capacity-forecast and understanding actual vs. planned hours. | EP | P2 |
Platform integrations to enhance
| Integration | Enhancement | Serves | Priority |
|---|---|---|---|
| Slack-to-Vault pipeline | Auto-ingest key Slack threads/channels into vault (like transcripts). CSO daily pulses, EP scope discussions, SL technical decisions all happen in Slack but are invisible to skills today. | All | P0 |
| Linear velocity dashboard | Surface per-person velocity, defect density, cycle time in The Forge. Feeds junior-velocity-tracker and client-health-report. | SL, EP | P1 |
| KPI automation | Track the Q1 KPIs automatically: communication consistency (CSO), workspace integrity (EP), ticket quality rate (SL). Surface in The Forge dashboard. | All | P1 |
| Standup transcript auto-tag | When a standup transcript arrives, auto-tag it by client and role segments (SL story, CSO pulse, EP guardrail, SL plan). Makes standup-recap much more reliable. | All | P1 |
| Google Slides template integration | Store Brainforge deck templates in Google Workspace so weekly-client-presentation and ebr-deck-generator can write to real Slides, not just Markdown. | CSO | P2 |
| SOW-to-Linear linkage | Store SOW objectives as structured data (not just a PDF) so value-realization-audit can map deliverables to contractual goals. | CSO, EP | P2 |
5. Skills by Role — “Out of the Box” Package
When an EP or CSO starts at Brainforge, they should have all of these available immediately.
CSO starter kit (7 skills)
| # | Skill | Cadence | What it replaces |
|---|---|---|---|
| 1 | client-pulse | Daily | Manual daily client Slack/email |
| 2 | standup-prep (CSO mode) | Daily | Winging the partnership pulse |
| 3 | standup-recap | Daily | Manual recap / AI recap without structure |
| 4 | weekly-client-presentation | Weekly | 90-min manual deck creation |
| 5 | client-health-report | Weekly | Manual KPI compilation |
| 6 | expansion-signal-scanner | Weekly | Gut-feel opportunity spotting |
| 7 | meeting-prep (existing) | As needed | Manual vault search |
| + | humanizer (existing) | As needed | Manual writing cleanup |
| + | value-realization-audit | Monthly | Manual SOW-vs-delivery comparison |
| + | ebr-deck-generator | Quarterly | Full-day EBR deck prep |
| + | retrospective-prep | Monthly | Unstructured retro prep |
EP starter kit (7 skills)
| # | Skill | Cadence | What it replaces |
|---|---|---|---|
| 1 | ticket-quality-audit | Daily | Manual board scanning |
| 2 | blocker-radar | Daily | Manual blocker hunting |
| 3 | standup-prep (EP mode) | Daily | Winging the guardrail check |
| 4 | standup-recap | Daily | Manual standup notes |
| 5 | client-health-report | Weekly | Manual health report |
| 6 | meeting-prep (existing) | As needed | Manual vault search |
| 7 | linear-audit (existing) | Daily | Manual Linear audit |
| + | humanizer (existing) | As needed | Manual writing cleanup |
| + | resource-capacity-forecast | Monthly | Spreadsheet capacity planning |
| + | retrospective-prep | Monthly | Unstructured retro prep |
SL starter kit (6 skills)
| # | Skill | Cadence | What it replaces |
|---|---|---|---|
| 1 | standup-prep (SL mode) | Daily | Winging the technical story |
| 2 | standup-recap | Daily | Manual standup notes |
| 3 | code-quality-digest | Daily/Weekly | Manual PR review summary |
| 4 | blocker-radar | Daily | Manual blocker hunting |
| 5 | meeting-prep (existing) | As needed | Manual vault search |
| 6 | junior-velocity-tracker | Monthly | Manual velocity spreadsheet |
| + | humanizer (existing) | As needed | Manual writing cleanup |
| + | asset-codification-scout | Monthly | Manual pattern hunting |
Shared across all roles
| Skill | Note |
|---|---|
meeting-prep | Already exists; enhance with Linear MCP integration |
humanizer | Already exists; no changes needed |
standup-prep | Multi-mode skill (CSO/EP/SL) |
standup-recap | Used by everyone post-standup |
6. Engineering Types — Skills Considerations
Engineering types (architects, analysts, developers, data engineers) are the IC craft underneath the leadership overlays. Skills for engineering types are distinct from leadership skills:
| Engineering Type | Relevant Leadership Overlay | Additional Skills to Consider |
|---|---|---|
| Developer (Frontend/Backend/Full-stack) | SL or EP | code-quality-digest, PR review helper, test coverage reporter |
| Data Analyst / Data Engineer | CSO or SL | client-health-report (data pull), dashboard generator, SQL helper |
| Architect | SL | asset-codification-scout, architecture decision record (ADR) generator |
| Designer | SL | Figma-to-spec skill (Figma MCP already exists), design QA checklist |
| AI/ML Engineer | SL | Agent/prompt template scaffolder, evaluation harness |
Engineering-type skills can be developed in a second phase after the leadership overlay skills are in place. They are closer to craft tools than management tools.
7. Implementation Roadmap
Phase 1 — Foundation (Weeks 1-3)
Goal: Ship the daily-cadence skills that every role uses.
| Week | Deliverable |
|---|---|
| 1 | standup-prep (all modes), standup-recap, GitHub MCP |
| 2 | client-pulse (CSO), ticket-quality-audit (EP), blocker-radar (EP+SL) |
| 3 | code-quality-digest (SL), Slack-to-Vault pipeline (platform) |
Phase 2 — Weekly Cadence (Weeks 4-6)
Goal: Ship the weekly skills that cut the most manual time.
| Week | Deliverable |
|---|---|
| 4 | weekly-client-presentation (CSO), client-health-report (CSO+EP) |
| 5 | expansion-signal-scanner (CSO), Instagant/Gantt MCP |
| 6 | KPI automation in The Forge, Linear velocity dashboard |
Phase 3 — Monthly/Quarterly + Polish (Weeks 7-9)
Goal: Ship the strategic and lifecycle skills.
| Week | Deliverable |
|---|---|
| 7 | value-realization-audit, retrospective-prep, resource-capacity-forecast |
| 8 | ebr-deck-generator, asset-codification-scout, junior-velocity-tracker |
| 9 | Google Slides integration, SOW-to-Linear linkage, Slack MCP, testing + polish |
Ongoing
- Collect feedback via the existing GTM agent feedback loop (adapt for delivery skills)
- Iterate on skill quality based on CSO/EP/SL real usage
- Expand to engineering-type craft skills in Phase 4
8. Success Metrics for the Platform Team
| Metric | Target | How to Measure |
|---|---|---|
| Skill adoption | 80% of CSOs/EPs/SLs use at least 3 skills weekly within 30 days of launch | Usage logging in Cursor (or self-report) |
| Time savings | 50% reduction in leadership overhead hours (from 8-10 hr/week to 4-5 hr/week) | Time tracking before/after |
| Standup quality | 100% of standups follow Heartbeat Protocol format | Transcript structure audit |
| Weekly presentation time | 90 min down to 20 min | Self-report |
| Linear audit coverage | 100% of clients audited daily | linear-audit run log |
| KPI tracking | All Q1 KPIs tracked automatically | Platform dashboard |
9. Cursor Rules to Ship Alongside Skills
Each skill should have a companion Cursor rule (.cursor/rules/) that triggers the skill contextually:
| Rule | Trigger | Skill |
|---|---|---|
client-pulse.mdc | ”draft client pulse”, “daily client message” | client-pulse |
standup-prep.mdc | ”prep for standup”, “standup notes” | standup-prep |
standup-recap.mdc | ”recap standup”, “standup summary” | standup-recap |
weekly-presentation.mdc | ”weekly presentation”, “client deck” | weekly-client-presentation |
health-report.mdc | ”health report”, “client health” | client-health-report |
code-digest.mdc | ”code digest”, “PR summary” | code-quality-digest |
blocker-radar.mdc | ”blockers”, “stalled tickets” | blocker-radar |
10. Summary
15 new skills to build (8 Tier 1, 7 Tier 2), adding to the 2 existing skills. 5 new MCPs needed (GitHub P0, Instagant P1, Slack P1, HubSpot P2, Clockify P2). 6 platform integrations to enhance. 9-week roadmap to ship the full package.
The outcome: when a new EP or CSO joins Brainforge, they open Cursor and have a complete toolkit for every task in their daily, weekly, and monthly cadence. No hunting for docs, no building their own automations. The delivery platform team builds it once; every leader benefits immediately.