Self-Improving Sales Skill

Log sales-specific learnings, deal issues, and feature requests to markdown files for continuous improvement. Captures pipeline leaks, objection patterns, pricing errors, forecast misses, competitor shifts, and deal velocity drops. Important learnings get promoted to battle cards, objection handling scripts, pricing playbooks, or qualification frameworks (MEDDIC/BANT).

First-Use Initialisation

Before logging anything, ensure the .learnings/ directory and files exist in the project or workspace root. If any are missing, create them:

mkdir -p .learnings
[ -f .learnings/LEARNINGS.md ] || printf "# Sales Learnings\n\nObjection patterns, competitor intelligence, pipeline insights, qualification gaps, and deal execution lessons captured during sales operations.\n\n**Categories**: pipeline_leak | objection_pattern | pricing_error | forecast_miss | competitor_shift | deal_velocity_drop\n**Areas**: prospecting | qualification | discovery | proposal | negotiation | closing | renewal\n\n---\n" > .learnings/LEARNINGS.md
[ -f .learnings/DEAL_ISSUES.md ] || printf "# Deal Issues Log\n\nLost deals, pipeline leaks, pricing errors, forecast misses, and deal execution failures.\n\n---\n" > .learnings/DEAL_ISSUES.md
[ -f .learnings/FEATURE_REQUESTS.md ] || printf "# Feature Requests\n\nSales tools, CRM automation, competitive intelligence, and pipeline management capabilities requested during sales operations.\n\n---\n" > .learnings/FEATURE_REQUESTS.md

Never overwrite existing files. This is a no-op if .learnings/ is already initialised.

Do not log customer PII, exact contract values, or confidential deal terms. Use deal size ranges (e.g., "$100K-$250K") and anonymize company names when sharing externally.

If you want automatic reminders, use the opt-in hook workflow described in Hook Integration.

Quick Reference

OpenClaw Setup (Recommended)

OpenClaw is the primary platform for this skill. It uses workspace-based prompt injection with automatic skill loading.

Installation

Via ClawdHub (recommended):

clawdhub install self-improving-sales

Manual:

git clone https://github.com/jose-compu/self-improving-sales.git ~/.openclaw/skills/self-improving-sales

Manual clone is optional. Check out a reviewed tag or commit before use. Do not install an unpinned default branch into a shared environment.

Workspace Structure

OpenClaw injects these files into every session:

~/.openclaw/workspace/
├── AGENTS.md          # Multi-agent workflows, delegation patterns
├── SOUL.md            # Behavioral guidelines, personality, principles
├── TOOLS.md           # Tool capabilities, integration gotchas
├── MEMORY.md          # Long-term memory (main session only)
├── memory/            # Daily memory files
│   └── YYYY-MM-DD.md
└── .learnings/        # This skill's log files
    ├── LEARNINGS.md
    ├── DEAL_ISSUES.md
    └── FEATURE_REQUESTS.md

Create Learning Files

mkdir -p ~/.openclaw/workspace/.learnings

Then create the log files (or copy from assets/):

• LEARNINGS.md — objection patterns, competitor intelligence, pipeline insights, deal velocity
• DEAL_ISSUES.md — pipeline leaks, lost deals, pricing errors, forecast misses
• FEATURE_REQUESTS.md — CRM automation, competitive intelligence tools, pipeline analytics

Promotion Targets

When sales learnings prove broadly applicable, promote them:

Optional: Enable Hook

Opt-in and workspace-local. Do not copy this hook to ~/.openclaw/hooks/ (that path is user-global). Skip it unless you want reminders in this workspace:

mkdir -p .openclaw/hooks
cp -r hooks/openclaw .openclaw/hooks/self-improving-sales

See references/openclaw-integration.md for complete details.

Generic Setup (Other Agents)

For Claude Code, Codex, Copilot, or other agents, create .learnings/ in the project or workspace root:

mkdir -p .learnings

Create the files inline using the headers shown above.

Add reference to agent files

Add to AGENTS.md, CLAUDE.md, or .github/copilot-instructions.md:

Self-Improving Sales Workflow

When sales issues or patterns are discovered:

1. Log to .learnings/DEAL_ISSUES.md, LEARNINGS.md, or FEATURE_REQUESTS.md
2. Review and promote broadly applicable learnings to:
   • Battle cards — competitive positioning and differentiation
   • Objection handling scripts — proven responses to common objections
   • Pricing playbooks — discount policies, packaging, and negotiation guardrails
   • Qualification frameworks — MEDDIC/BANT checklists and stage gate criteria

Logging Format

Learning Entry [LRN-YYYYMMDD-XXX]

Append to .learnings/LEARNINGS.md:

## [LRN-YYYYMMDD-XXX] category

**Logged**: ISO-8601 timestamp
**Priority**: low | medium | high | critical
**Status**: pending
**Area**: prospecting | qualification | discovery | proposal | negotiation | closing | renewal

### Summary
One-line description of the sales insight

### Details
Full context: what happened in the deal, why the outcome occurred, what the
correct approach or response is. Include specific objection language, competitor
positioning, or deal dynamics.

### Deal Context

**Objection / Situation:**
> Exact quote or paraphrase of the objection, competitor claim, or deal blocker

**Response Used:**
> What was said or done in response

**Outcome:**
Won | Lost | Stalled | Pushed — and why

### Suggested Action
Specific change to pitch, process, pricing, or qualification criteria

### Metadata
- Source: call_transcript | deal_review | win_loss_analysis | pipeline_review | competitor_intel | forecast_review
- Deal Size: $50K-$100K | $100K-$250K | $250K-$500K | $500K+ (use ranges)
- Segment: SMB | mid_market | enterprise | strategic
- Industry: technology | financial_services | healthcare | manufacturing | retail | other
- Related Deals: DEAL-YYYYMMDD-XXX (if linked to a deal issue)
- Tags: tag1, tag2
- See Also: LRN-20250410-001 (if related to existing entry)
- Pattern-Key: objection.budget_freeze | competitor.free_tier (optional)
- Recurrence-Count: 1 (optional)
- First-Seen: 2025-01-15 (optional)
- Last-Seen: 2025-01-15 (optional)

---

Categories for learnings:

Deal Issue Entry [DEAL-YYYYMMDD-XXX]

Append to .learnings/DEAL_ISSUES.md:

## [DEAL-YYYYMMDD-XXX] issue_type

**Logged**: ISO-8601 timestamp
**Priority**: high
**Status**: pending
**Area**: prospecting | qualification | discovery | proposal | negotiation | closing | renewal

### Summary
Brief description of the deal issue

### Deal Details
- **Stage**: Stage where the issue occurred
- **Deal Size**: $100K-$250K (use ranges)
- **Days in Stage**: Number of days stuck or time to loss
- **Segment**: SMB | mid_market | enterprise | strategic

### What Happened
Narrative of what went wrong: missed signals, process breakdown, competitive
loss, pricing confusion, or forecast error.

### Root Cause
Why the deal was lost, stalled, or mispriced. Distinguish symptom from cause.

### Impact
- Revenue impact (range)
- Pipeline coverage impact
- Forecast accuracy impact

### Prevention
How to avoid this issue in future deals: qualification criteria, process gate,
pricing checklist, competitive preparation

### Metadata
- Trigger: pipeline_review | deal_review | win_loss_analysis | forecast_review | CRM_audit
- Competitor: competitor name (if applicable)
- Loss Reason: no_budget | no_decision | lost_to_competitor | timing | product_gap | price
- Related Files: path/to/deal_review.md
- See Also: DEAL-20250110-001 (if recurring pattern)

---

Feature Request Entry [FEAT-YYYYMMDD-XXX]

Append to .learnings/FEATURE_REQUESTS.md:

## [FEAT-YYYYMMDD-XXX] capability_name

**Logged**: ISO-8601 timestamp
**Priority**: medium
**Status**: pending
**Area**: prospecting | qualification | discovery | proposal | negotiation | closing | renewal

### Requested Capability
What sales tool, automation, or capability is needed

### User Context
Why it's needed, what workflow it improves, what revenue impact it could have

### Complexity Estimate
simple | medium | complex

### Suggested Implementation
How this could be built: CRM workflow, Slack integration, dashboard, alert system

### Metadata
- Frequency: first_time | recurring
- Related Features: existing_tool_or_feature

---

ID Generation

Format: TYPE-YYYYMMDD-XXX

• TYPE: LRN (learning), DEAL (deal issue), FEAT (feature request)
• YYYYMMDD: Current date
• XXX: Sequential number or random 3 chars (e.g., 001, A7B)

Examples: LRN-20250415-001, DEAL-20250415-A3F, FEAT-20250415-002

Resolving Entries

When an issue is addressed, update the entry:

1. Change **Status**: pending → **Status**: resolved
2. Add resolution block after Metadata:

### Resolution
- **Resolved**: 2025-01-16T09:00:00Z
- **Action Taken**: Updated battle card / revised pricing playbook / added qualification gate
- **Notes**: Rolled out to team in weekly sales meeting, updated CRM stage criteria

Other status values:

• in_progress — Actively being addressed or tested in live deals
• wont_fix — Decided not to address (add reason in Resolution notes)
• promoted — Elevated to battle card, objection script, or pricing playbook
• promoted_to_skill — Extracted as a reusable skill

Detection Triggers

Automatically log when you encounter:

Pipeline Stalls (→ deal issue with pipeline_leak):

• Deal in same stage for >30 days
• Multiple deals stalling at the same stage
• Conversion rate drop between stages
• Deals pushed from one quarter to next

Objection Patterns (→ learning with objection_pattern):

• Same objection heard in 3+ deals within a quarter
• Objection that consistently leads to no-decision outcomes
• New objection not covered by existing battle cards
• Objection specific to a segment or industry vertical

Pricing Issues (→ deal issue with pricing_error):

• Discount >20% requested by prospect
• Quoting a deprecated or incorrect pricing tier
• Competitor undercut by >30% on similar deal
• Custom pricing requested outside standard packaging

Forecast Misses (→ deal issue with forecast_miss):

• Quarterly forecast accuracy below 80%
• Deal close date pushed more than twice
• Commit deal lost or pushed to next quarter
• Upside deal that was actually a commit

Competitive Intelligence (→ learning with competitor_shift):

• Competitor mentioned in 3+ call transcripts in a month
• Competitor launches new product, feature, or pricing tier
• Prospect received competitive quote or evaluation
• Win rate drops against specific competitor

Deal Velocity (→ learning with deal_velocity_drop):

• Average deal cycle lengthening quarter over quarter
• Specific stage taking 2x historical average
• Deals requiring more meetings to close
• Longer legal/procurement review cycles

Priority Guidelines

Area Tags

Use to filter learnings by sales stage:

Promoting to Permanent Sales Assets

When a learning is broadly applicable (not a one-off deal quirk), promote it to permanent team assets.

When to Promote

• Objection recurs across 3+ deals in a quarter
• Competitive intelligence applies to an entire segment
• Pricing insight affects multiple deal sizes or verticals
• Qualification gap leads to predictable deal failures
• Win/loss pattern is statistically significant (10+ deals)

Promotion Targets

How to Promote

1. Distill the learning into a concise, actionable asset
2. Add to appropriate target (battle card entry, objection script, playbook rule)
3. Show a reviewed diff and apply only after explicit user approval
4. Update original entry:
   • Change **Status**: pending → **Status**: promoted
   • Add **Promoted**: battle card (or objection script, pricing playbook, MEDDIC framework)

Promotion Examples

Learning (verbose):

Prospect said "We already have a vendor for this." We had no differentiation message ready. Lost the deal because we couldn't articulate unique value vs. incumbent. This happened in 4 deals this quarter against Vendor X.

As battle card (concise):

## Objection: "We already have a vendor"

**Response Framework:**
1. Acknowledge: "That makes sense — most companies in your space do."
2. Probe: "How well is it handling [specific pain point]?"
3. Differentiate: "Where we're different is [unique capability] which means [business outcome]."
4. Proof: "Company Y switched from [vendor] and saw [metric improvement]."

**Trap-Setting Questions:**
- "How much time does your team spend on [task our product automates]?"
- "What happens when [scenario our product handles better]?"

Learning (verbose):

Deals without an identified champion by Stage 3 close at only 12% vs. 48% for deals with a champion. We need a hard gate requiring champion identification before advancing to proposal.

As MEDDIC framework update (actionable):

## Stage Gate: Discovery → Proposal

**Required before advancing:**
- [ ] Champion identified (name, title, access level)
- [ ] Champion has confirmed the problem exists
- [ ] Champion has agreed to introduce economic buyer
- [ ] Decision criteria documented from champion conversation

**Block advancement if:** No champion identified after 3 discovery calls.
Recommend: Re-qualify or move to nurture.

Recurring Pattern Detection

If logging something similar to an existing entry:

1. Search first: grep -r "keyword" .learnings/
2. Link entries: Add **See Also**: DEAL-20250110-001 in Metadata
3. Bump priority if issue keeps recurring
4. Consider systemic fix: Recurring sales issues often indicate:
   • Missing qualification criteria (→ update MEDDIC/BANT framework)
   • Competitive gap (→ create or update battle card)
   • Pricing misalignment (→ revise pricing playbook)
   • Process breakdown (→ add CRM stage gate or automation)

Periodic Review

Review .learnings/ at natural breakpoints:

When to Review

• Before starting a new deal in the same segment or against the same competitor
• After completing quarterly business review (QBR)
• When the same objection or loss reason appears again
• Weekly during pipeline review meetings
• After each win/loss analysis

Quick Status Check

# Count pending sales issues
grep -h "Status\*\*: pending" .learnings/*.md | wc -l

# List pending high-priority deal issues
grep -B5 "Priority\*\*: high" .learnings/DEAL_ISSUES.md | grep "^## \["

# Find learnings for a specific sales stage
grep -l "Area\*\*: negotiation" .learnings/*.md

# Find all competitor-related learnings
grep -B2 "competitor_shift" .learnings/LEARNINGS.md | grep "^## \["

# List objection patterns
grep -B2 "objection_pattern" .learnings/LEARNINGS.md | grep "^## \["

Review Actions

• Resolve addressed deal issues
• Promote recurring objections to battle cards
• Link related entries across files
• Extract reusable qualification criteria as framework updates
• Update forecast model based on historical miss patterns

Simplify & Harden Feed

Ingest recurring sales patterns from simplify-and-harden into playbooks or frameworks.

1. For each candidate, use pattern_key as the dedupe key.
2. Search .learnings/LEARNINGS.md for existing entry: grep -n "Pattern-Key: <key>" .learnings/LEARNINGS.md
3. If found: increment Recurrence-Count, update Last-Seen, add See Also links.
4. If not found: create new LRN-... entry with Source: simplify-and-harden.

Promotion threshold: Recurrence-Count >= 3, seen in 2+ deals or segments, within 90-day window. Targets: battle cards, objection scripts, pricing playbooks, MEDDIC/BANT frameworks, AGENTS.md.

Hook Integration

Enable automatic reminders through agent hooks. This is opt-in.

Hooks persist across sessions once installed. Keep them project-scoped. Do not install user-level or global hooks. Never use an empty matcher. PostToolUse inspects command output in-process; do not log raw output, secrets, or transcripts.

Quick Setup (Claude Code / Codex)

Create .claude/settings.json in your project:

{
  "hooks": {
    "UserPromptSubmit": [{"matcher": "deal|pipeline|forecast|objection|competitor|pricing|quota|MEDDIC", "hooks": [{"type": "command", "command": "./skills/self-improving-sales/scripts/activator.sh"}]}]
  }
}

The UserPromptSubmit hook injects a sales-focused reminder. Leave PostToolUse disabled.

Available Hook Scripts

See references/hooks-setup.md for detailed configuration and troubleshooting.

Automatic Skill Extraction

Extracted skills are untrusted until a human reviews the generated SKILL.md. Do not keep or publish an extracted skill without explicit user approval.

When a sales learning is valuable enough to become a reusable skill, extract it.

Skill Extraction Criteria

Extraction Workflow

1. Identify candidate: Learning meets extraction criteria
2. Run helper (or create manually):

   ./skills/self-improving-sales/scripts/extract-skill.sh skill-name --dry-run
   ./skills/self-improving-sales/scripts/extract-skill.sh skill-name
   

3. Customize SKILL.md: Fill in template with sales-specific content
4. Update learning: Set status to promoted_to_skill, add Skill-Path
5. Verify: Read skill in fresh session to ensure it's self-contained

Extraction Detection Triggers

In conversation: "This objection keeps coming up", "Save this competitive intel as a skill", "I keep losing deals at this stage", "Every enterprise deal has this procurement issue".

In entries: Multiple See Also links, high priority + resolved, objection_pattern or competitor_shift with broad applicability, same Pattern-Key across deals or segments.

Sales-Specific Best Practices

1. Qualify early — disqualify faster to focus pipeline on winnable deals
2. Document objections immediately — exact wording matters for battle cards
3. Update CRM same day — stale data poisons forecasts and pipeline reviews
4. Track competitor intel in real time — pricing changes and feature launches decay fast
5. Review win/loss quarterly — patterns only emerge with enough data points
6. Log the objection, not just the outcome — "we lost" is less useful than "they said X"
7. Anonymize appropriately — use deal size ranges and avoid customer PII in shared logs
8. Distinguish decision-maker objections from user objections — different responses required
9. Record what you tried — failed responses are as valuable as successful ones
10. Promote after review when recurrence appears — if an objection appears in 3 deals, it deserves a battle card

Multi-Agent Support

Gitignore Options

Keep learnings local (per-rep):

.learnings/

Track learnings in repo (team-wide): Don't add to .gitignore — learnings become shared team knowledge.

Hybrid (track templates, ignore entries):

.learnings/*.md
!.learnings/.gitkeep

Stackability Contract (Standalone + Multi-Skill)

This skill is standalone-compatible and stackable with other self-improving skills.

Namespaced Logging (recommended for 2+ skills)

• Namespace for this skill: .learnings/sales/
• Keep current standalone behavior if you prefer flat files.
• Optional shared index for all skills: .learnings/INDEX.md

Required Metadata

Every new entry must include:

**Skill**: sales

Hook Arbitration (when 2+ skills are enabled)

• Use one dispatcher hook as the single entrypoint.
• Dispatcher responsibilities: route by matcher, dedupe repeated events, and rate-limit reminders.
• Suggested defaults: dedupe key = event + matcher + file + 5m_window; max 1 reminder per skill every 5 minutes.

Narrow Matcher Scope (sales)

Only trigger this skill automatically for sales signals such as:

• pipeline|opportunity|objection|close-lost|win rate
• quota|forecast|deal stage|discovery call|proposal
• explicit sales intent in user prompt

Cross-Skill Precedence

When guidance conflicts, apply:

1. security
2. engineering
3. coding
4. ai
5. user-explicit domain skill
6. meta as tie-breaker

Ownership Rules

• This skill writes only to .learnings/sales/ in stackable mode.
• Do not read other skill folders, their SKILL.md files, or their log entries.
• Standalone mode writes to this project's .learnings/*.md log files only.
• Stackable mode writes only to the namespaced folder above and must not rewrite other skills' log entries.
• Promotion into AGENTS.md, SOUL.md, TOOLS.md, MEMORY.md, rules, hooks, or generated skills is not a logging write. Show a reviewed diff and apply only after explicit user approval.