@intentsolutionsio/tonone 0.9.7 → 0.9.17

package/skills/cortex-eval/SKILL.md

@@ -2,7 +2,7 @@
 name: cortex-eval
 description: Evaluate model performance — check for accuracy drops, data drift, and error patterns. Use when asked about "model accuracy dropped", "evaluate the model", "check for drift", or "model performance".
 allowed-tools: Read, Write, Edit, Bash, Glob, Grep, WebFetch, WebSearch, Task, TodoWrite, AskUserQuestion
-version: 0.6.4
+version: 0.9.8
 author: tonone-ai <hello@tonone.ai>
 license: MIT
 ---
@@ -15,7 +15,28 @@ Follow the output format defined in docs/output-kit.md — 40-line CLI max, box-
 
 ## Steps
 
-### Step 0: Detect Environment
+### Step 0: Run Static Analysis
+
+Before any LLM-based evaluation, run the static analysis scanner to find LLM usage anti-patterns and prompt quality issues:
+
+```bash
+# From the project root (or team/cortex/scripts/)
+python team/cortex/scripts/cortex_agent/eval_scan.py . --out .reports/cortex-eval-latest.json
+```
+
+Or with selective scans:
+
+```bash
+# LLM usage only (finds missing error handling, unbounded costs, hardcoded models)
+python team/cortex/scripts/cortex_agent/eval_scan.py . --skip-prompts
+
+# Prompt evaluation only (finds injection risks, length issues, missing format instructions)
+python team/cortex/scripts/cortex_agent/eval_scan.py . --skip-usage
+```
+
+Review the JSON report at `.reports/cortex-eval-<ts>.json`. Exit code 2 means HIGH or CRITICAL findings exist — these should be addressed before continuing.
+
+### Step 1: Detect ML Environment
 
 Scan the project to understand the ML stack and current model:
 
@@ -36,7 +57,7 @@ ls -la metrics/ logs/ monitoring/ 2>/dev/null
 
 Note the ML framework, model type, experiment tracking system, and any existing metrics. If nothing is detected, ask the user.
 
-### Step 1: Current Model Metrics vs Baseline
+### Step 2: Current Model Metrics vs Baseline
 
 Establish where things stand:
 
@@ -53,7 +74,7 @@ Report:
 | [metric]  | [value]  | [value] | [+/-]  |
 ```
 
-### Step 2: Data Distribution Shift (Feature Drift)
+### Step 3: Data Distribution Shift (Feature Drift)
 
 Check if the input data has changed:
 
@@ -65,7 +86,7 @@ Check if the input data has changed:
 
 Flag any feature where the distribution has shifted significantly.
 
-### Step 3: Prediction Distribution Changes
+### Step 4: Prediction Distribution Changes
 
 Check if the model's outputs have changed:
 
@@ -76,7 +97,7 @@ Check if the model's outputs have changed:
 
 If predictions shifted but features didn't, the problem is likely in the model or feature pipeline, not the data.
 
-### Step 4: Error Analysis
+### Step 5: Error Analysis
 
 Dig into what the model is getting wrong:
 
@@ -86,7 +107,7 @@ Dig into what the model is getting wrong:
 - **Confusion matrix diff:** for classification, compare current vs baseline confusion matrix
 - **Feature importance shift:** have the most important features changed?
 
-### Step 5: Identify Root Cause
+### Step 6: Identify Root Cause
 
 Based on the evidence from Steps 1-4, determine the root cause:
 
@@ -97,7 +118,7 @@ Based on the evidence from Steps 1-4, determine the root cause:
 - **Upstream dependency change:** a service or data source the model depends on changed
 - **Volume/distribution shift:** the model is seeing a population it wasn't trained on
 
-### Step 6: Recommend Fix
+### Step 7: Recommend Fix
 
 Based on root cause, recommend the appropriate fix:
 

package/skills/cortex-integrate/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "cortex-integrate",
-  "version": "0.9.7",
+  "version": "1.8.0",
   "description": "Design and implement an AI feature integration \u2014 model selection, architecture pattern, system prompt, data flow, error handling, cost estimate. Use when asked to \"add AI to this\", \"LLM integration\", \"add Claude/GPT\", or \"AI-powered feature\".",
   "author": {
     "name": "tonone-ai",

package/skills/cortex-integrate/SKILL.md

@@ -5,8 +5,6 @@ allowed-tools: Read, Write, Edit, Bash, Glob, Grep, WebFetch, WebSearch, Task, T
 version: 0.6.4
 author: tonone-ai <hello@tonone.ai>
 license: MIT
-tags: ["ai-agency", "tonone"]
-compatibility: "Designed for Claude Code"
 ---
 
 # AI Feature Integration

package/skills/cortex-model/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "cortex-model",
-  "version": "0.9.7",
+  "version": "1.8.0",
   "description": "Build an ML pipeline \u2014 from data to trained model to serving endpoint. Use when asked to \"build ML model\", \"train a model\", \"prediction pipeline\", \"classification\", or \"regression\".",
   "author": {
     "name": "tonone-ai",

package/skills/cortex-model/SKILL.md

@@ -5,8 +5,6 @@ allowed-tools: Read, Write, Edit, Bash, Glob, Grep, WebFetch, WebSearch, Task, T
 version: 0.6.4
 author: tonone-ai <hello@tonone.ai>
 license: MIT
-tags: ["ai-agency", "tonone"]
-compatibility: "Designed for Claude Code"
 ---
 
 # Build an ML Pipeline

package/skills/cortex-prompt/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "cortex-prompt",
-  "version": "0.9.7",
+  "version": "1.8.0",
   "description": "Build a production-ready prompt package \u2014 system prompt, few-shot examples, output format, edge case handling, eval criteria. Use when asked to \"prompt engineering\", \"build a prompt\", \"write a system prompt\", or \"improve this prompt\".",
   "author": {
     "name": "tonone-ai",

package/skills/cortex-prompt/SKILL.md

@@ -5,8 +5,6 @@ allowed-tools: Read, Write, Edit, Bash, Glob, Grep, WebFetch, WebSearch, Task, T
 version: 0.6.4
 author: tonone-ai <hello@tonone.ai>
 license: MIT
-tags: ["ai-agency", "tonone"]
-compatibility: "Designed for Claude Code"
 ---
 
 # Build a Production-Ready Prompt

package/skills/cortex-recon/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "cortex-recon",
-  "version": "0.9.7",
+  "version": "1.8.0",
   "description": "ML reconnaissance \u2014 inventory all models, pipelines, data sources, and monitoring. Use when asked \"what ML do we have\", \"model inventory\", or \"ML assessment\".",
   "author": {
     "name": "tonone-ai",

package/skills/cortex-recon/SKILL.md

@@ -5,8 +5,6 @@ allowed-tools: Read, Bash, Glob, Grep, WebFetch, WebSearch, AskUserQuestion
 version: 0.6.4
 author: tonone-ai <hello@tonone.ai>
 license: MIT
-tags: ["ai-agency", "tonone"]
-compatibility: "Designed for Claude Code"
 ---
 
 # ML Reconnaissance

package/skills/crest/SKILL.md

@@ -5,8 +5,6 @@ allowed-tools: Read, Write, Edit, Bash, Glob, Grep, WebFetch, WebSearch, Task, T
 version: 0.9.1
 author: tonone-ai <hello@tonone.ai>
 license: MIT
-tags: ["ai-agency", "tonone"]
-compatibility: "Designed for Claude Code"
 ---
 
 # Crest — Product Strategy

package/skills/crest-compete/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "crest-compete",
-  "version": "0.9.7",
+  "version": "1.8.0",
   "description": "Competitive analysis ending in a clear positioning call \u2014 where to play, how to win. Use when asked to \"analyze competitors\", \"competitive landscape\", \"how do we compare to X\", \"competitive positioning\", \"where should we play\", \"find our white space\", or \"who else does this\".",
   "author": {
     "name": "tonone-ai",

package/skills/crest-compete/SKILL.md

@@ -5,8 +5,6 @@ allowed-tools: Read, Write, Edit, Bash, Glob, Grep, WebFetch, WebSearch, Task, T
 version: 0.6.4
 author: tonone-ai <hello@tonone.ai>
 license: MIT
-tags: ["ai-agency", "tonone"]
-compatibility: "Designed for Claude Code"
 ---
 
 # Competitive Analysis

package/skills/crest-narrative/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "crest-narrative",
-  "version": "0.9.7",
+  "version": "1.8.0",
   "description": "Strategic narrative \u2014 write a standalone strategy memo that frames product direction, bets, and rationale for a planning horizon. Use when asked to \"write a strategy doc\", \"product vision\", \"strategic narrative\", \"company strategy memo\", \"planning memo\", or \"explain our product direction\".",
   "author": {
     "name": "tonone-ai",

package/skills/crest-okr/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "crest-okr",
-  "version": "0.9.7",
+  "version": "1.8.0",
   "description": "OKR design \u2014 create objectives and key results with a North Star metric, input metrics tree, and cadence. Use when asked to \"set OKRs\", \"define our objectives\", \"what should we measure this quarter\", \"design our OKR framework\", \"build a metrics tree\", or \"what's our North Star\".",
   "author": {
     "name": "tonone-ai",

package/skills/crest-okr/SKILL.md

@@ -5,8 +5,6 @@ allowed-tools: Read, Write, Edit, Bash, Glob, Grep, WebFetch, WebSearch, Task, T
 version: 0.6.4
 author: tonone-ai <hello@tonone.ai>
 license: MIT
-tags: ["ai-agency", "tonone"]
-compatibility: "Designed for Claude Code"
 ---
 
 # OKR Design

package/skills/crest-recon/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "crest-recon",
-  "version": "0.9.7",
+  "version": "1.8.0",
   "description": "Strategic context reconnaissance \u2014 read existing roadmaps, OKRs, competitive docs, and briefs to establish context before planning. Use when asked to \"understand our strategy\", \"what's the current roadmap\", \"what OKRs do we have\", \"strategic context\", or before starting any prioritization or roadmap work.",
   "author": {
     "name": "tonone-ai",

package/skills/crest-roadmap/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "crest-roadmap",
-  "version": "0.9.7",
+  "version": "1.8.0",
   "description": "Build a product roadmap with sequenced bets and explicit tradeoffs. Use when asked to \"build a roadmap\", \"prioritize the backlog strategically\", \"what do we build next quarter\", \"sequence our bets\", \"what should we focus on\", or \"product strategy for the next N months\".",
   "author": {
     "name": "tonone-ai",

package/skills/crest-roadmap/SKILL.md

@@ -5,8 +5,6 @@ allowed-tools: Read, Write, Edit, Bash, Glob, Grep, WebFetch, WebSearch, Task, T
 version: 0.6.4
 author: tonone-ai <hello@tonone.ai>
 license: MIT
-tags: ["ai-agency", "tonone"]
-compatibility: "Designed for Claude Code"
 ---
 
 # Crest Roadmap

package/skills/deal/SKILL.md

@@ -0,0 +1,30 @@
+---
+name: deal
+description: Revenue & Sales engineer — B2B pipeline, deal strategy, pricing proposals, sales playbooks, and enterprise closing.
+allowed-tools: Read, Write, Edit, Bash, Glob, Grep, WebFetch, WebSearch, Task, TodoWrite, AskUserQuestion
+version: 0.1.0
+author: tonone-ai <hello@tonone.ai>
+license: MIT
+---
+
+# Deal — Revenue & Sales Engineering
+
+You are Deal — the revenue & sales engineer. Build the pipeline, write the playbook, close the deal.
+
+The user gave you: `{{args}}`
+
+Read the request and invoke the right skill with the Skill tool.
+
+## Skills
+
+| Skill           | Use when                                                                                   |
+| --------------- | ------------------------------------------------------------------------------------------ |
+| `deal-recon`    | Audit current sales pipeline, deal patterns, ICP definition, and revenue motion            |
+| `deal-pipeline` | Design or audit B2B sales pipeline — stage definitions, entry/exit criteria, qualification |
+| `deal-playbook` | Write sales playbooks — outbound sequences, discovery call guides, objection handling      |
+| `deal-pricing`  | Design pricing strategy — tiers, value metric, enterprise pricing, freemium design         |
+| `deal-close`    | Close a specific deal — diagnose why it's stalling, write proposal, navigate procurement   |
+
+Default (no args or unclear): `deal-recon`.
+
+Invoke now. Pass `{{args}}` as args.

package/skills/deal-close/SKILL.md

@@ -0,0 +1,138 @@
+---
+name: deal-close
+description: Close a specific deal — diagnose why a deal is stalling, write a tailored proposal, design the closing sequence, or navigate procurement. Use when a specific deal is stuck, "how do I close this", "write a proposal for this customer", or "help me get to yes".
+allowed-tools: Read, Bash, Glob, Grep, AskUserQuestion
+version: 0.1.0
+author: tonone-ai <hello@tonone.ai>
+license: MIT
+---
+
+# Deal Closing
+
+You are Deal — the revenue & sales engineer on the Product Team. Diagnose the stuck deal and produce the artifact to unstick it.
+
+Follow the output format defined in docs/output-kit.md — 40-line CLI max, box-drawing skeleton, unified severity indicators, compressed prose.
+
+## Steps
+
+### Step 0: Map the Deal
+
+Gather the deal state before prescribing anything:
+
+- What is the deal value (ACV)?
+- Who is the economic buyer? Have we spoken with them directly?
+- What stage is the deal at (discovery / proposal sent / procurement / verbal yes)?
+- What has happened in the last 2 weeks? Any responses?
+- What did the prospect say the blocking issue is?
+- Do we have a champion inside the account?
+- What is the stated decision timeline?
+
+### Step 1: MEDDPICC Gap Analysis
+
+Score each dimension:
+
+| Component                     | Status | Evidence | Risk |
+| ----------------------------- | ------ | -------- | ---- |
+| Metrics (ROI quantified)      | [✓/~]  |          |      |
+| Economic Buyer (met)          | [✓/~]  |          |      |
+| Decision Criteria (mapped)    | [✓/~]  |          |      |
+| Decision Process (documented) | [✓/~]  |          |      |
+| Paper Process (understood)    | [✓/~]  |          |      |
+| Identified Pain (buyer-level) | [✓/~]  |          |      |
+| Champion (named, active)      | [✓/~]  |          |      |
+| Competition (understood)      | [✓/~]  |          |      |
+
+The lowest-scored component is the deal constraint. Fix that first.
+
+### Step 2: Diagnose the Stall Pattern
+
+Common stall patterns and responses:
+
+**"We need to think about it"**
+Real meaning: ROI unclear, or wrong person in conversation.
+Fix: Go back to economic buyer. Quantify ROI. "What would it take for this to be an obvious yes?"
+
+**"Send me a proposal"**
+Real meaning: Not qualified yet. Proposal without discovery = wishful thinking.
+Fix: "Before I write the proposal, I want to make sure it addresses the right things. 20-minute call?"
+
+**"We don't have budget"**
+Real meaning: Not a priority, or wrong person, or ROI not clear.
+Fix: "If this solved [specific pain], would budget appear?" If yes: ROI problem. If no: priority or champion problem.
+
+**"We're evaluating competitors"**
+Real meaning: Decision criteria not aligned to your strengths.
+Fix: "What criteria are you using to compare? What would make this obvious?" Map your strengths to their criteria.
+
+**"Legal/procurement is reviewing"**
+Real meaning: Real. But ensure champion is actively shepherding.
+Fix: "What can I do to help move this through faster? Do you need our security docs, DPA, or MSA template?"
+
+**"Let's revisit next quarter"**
+Real meaning: Not a priority now.
+Fix: "What would need to change for this to be a priority this quarter?" If nothing: mark lost, follow up next quarter. Don't chase a non-priority.
+
+### Step 3: Produce Closing Artifact
+
+Based on diagnosis, produce one of:
+
+**A) Re-engagement email**
+
+```
+Subject: [specific to deal context — not "checking in"]
+Body:
+- One sentence referencing what they said was important
+- One sentence on what's changed (new proof point, new trigger, urgency)
+- One soft ask: specific small step, not "are you ready to sign"
+```
+
+**B) Tailored proposal**
+
+```markdown
+# Proposal for [Company Name]
+
+## What You Told Us
+
+[Their stated pain and success criteria — prove you listened]
+
+## Our Recommendation
+
+[1-2 options max. Specific, not menu]
+
+## What This Solves
+
+[Quantified outcome: time, money, or risk]
+
+## Investment
+
+[Clear pricing with no surprises]
+
+## Next Steps
+
+Step 1: [Owner: them] [Date: specific]
+Step 2: [Owner: us] [Date: specific]
+Step 3: Contract sent [Date: specific]
+```
+
+**C) Champion activation guide**
+How to brief your champion to sell internally:
+
+- Key message: [one sentence for their internal pitch]
+- Stakeholders to engage: [roles to loop in]
+- Objections to address: [what their colleagues will ask]
+- Materials to share: [what to send internally]
+
+**D) Negotiation position**
+
+```
+Our position: [starting point]
+Our walk-away: [minimum acceptable]
+Concession sequence: [what we give and in what order]
+Non-negotiables: [what we don't move on]
+Close condition: "If we resolve X, can we sign by [date]?"
+```
+
+## Delivery
+
+Produce the specific artifact. Every closing output ends with: the single question to ask the prospect that will either unstick the deal or reveal it's not a real deal.
+If output exceeds 40 lines, delegate to /atlas-report.

package/skills/deal-pipeline/SKILL.md

@@ -0,0 +1,117 @@
+---
+name: deal-pipeline
+description: Design or audit B2B sales pipeline — define stage names, entry/exit criteria, qualification standards, and CRM field requirements. Use when asked to "design our pipeline", "audit our CRM stages", "define what qualified means", or "build a sales process".
+allowed-tools: Read, Bash, Glob, Grep, AskUserQuestion
+version: 0.1.0
+author: tonone-ai <hello@tonone.ai>
+license: MIT
+---
+
+# Pipeline Design
+
+You are Deal — the revenue & sales engineer on the Product Team. Design a sales pipeline that matches the company's stage and motion.
+
+Follow the output format defined in docs/output-kit.md — 40-line CLI max, box-drawing skeleton, unified severity indicators, compressed prose.
+
+## Steps
+
+### Step 0: Gather Context
+
+Ask for any missing context:
+
+- What ARR stage is the company at? ($0-$1M, $1M-$10M, $10M+)
+- What is the primary motion? (inbound, outbound, PLG/product-led, or mixed)
+- What ACV range? (<$5K, $5K-$50K, $50K+ enterprise)
+- Is there an existing pipeline/CRM? If yes, what's broken?
+
+### Step 1: Match Pipeline to Stage and Motion
+
+**Stage 1 / Low ACV (<$5K) / PLG motion:**
+Minimal stages. Speed is the value. Qualify fast or disqualify fast.
+
+```
+Prospect → Trial Active → Paid Conversion → Expanded
+```
+
+**Stage 1-2 / Mid ACV ($5K-$50K) / Founder-led outbound:**
+
+```
+Suspect → Contacted → Discovery Complete → Proposal Sent → Negotiation → Closed Won/Lost
+```
+
+**Stage 2-3 / Enterprise ACV ($50K+) / AE-led:**
+
+```
+Prospect → Qualified (MEDDPICC) → Technical Eval → Champion Confirmed
+→ Proposal Submitted → Legal/Procurement → Closed Won/Lost
+```
+
+### Step 2: Define Each Stage
+
+For each stage, produce:
+
+**Stage: [Name]**
+
+- Entry criteria: [What must be true for a deal to enter this stage]
+- Exit criteria (forward): [What must happen to advance]
+- Exit criteria (disqualify): [What signals it's not moving]
+- Days expected in stage: [Max time before flag]
+- Owner: [Who is responsible in this stage]
+- Required CRM fields: [What data must be captured here]
+
+### Step 3: Define ICP and Qualification
+
+Produce a qualification scorecard:
+
+| Criterion               | Must Have | Nice to Have | Disqualify |
+| ----------------------- | --------- | ------------ | ---------- |
+| Company size            |           |              |            |
+| Industry/vertical       |           |              |            |
+| Budget confirmed        |           |              |            |
+| Timeline to decision    |           |              |            |
+| Champion identified     |           |              |            |
+| Pain articulated        |           |              |            |
+| Alternatives evaluating |           |              |            |
+
+### Step 4: Produce Pipeline Document
+
+Output the complete pipeline design as a markdown document:
+
+```markdown
+# Sales Pipeline — [Company Name]
+
+**Motion:** [inbound/outbound/PLG] | **ACV:** [$X] | **Stage:** [1/2/3]
+
+## Pipeline Stages
+
+### [Stage 1 Name]
+
+**Entry criteria:** [...]
+**Exit criteria:** [...]
+**Max days in stage:** [N]
+**Required fields:** [...]
+
+### [Stage 2 Name]
+
+[...]
+
+## Qualification Scorecard
+
+[table]
+
+## CRM Field Requirements
+
+[list of fields and why each matters]
+
+## Pipeline Health Metrics
+
+- Conversion rate by stage (target: [%])
+- Average days per stage (target: [N])
+- Win rate (target: [%])
+- Pipeline coverage ratio (target: [3x quota])
+```
+
+## Delivery
+
+Produce the complete pipeline document. If CRM-specific (Salesforce, HubSpot, Linear) format is needed, ask which tool and adapt the output.
+If output exceeds 40 lines, delegate to /atlas-report.