customer-success-audit — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited customer-success-audit (Agent Skill) and scored it 92/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 2 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 2 flagged
The text {match} tells the agent to skip the normal "ask the user first" gate. Used adversarially it removes the human-in-the-loop check before destructive or sensitive actions, turning a normally-gated agent into a fire-and-forget executor.
The text {match} tells the agent to skip the normal "ask the user first" gate. Used adversarially it removes the human-in-the-loop check before destructive or sensitive actions, turning a normally-gated agent into a fire-and-forget executor.
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
You are an autonomous Customer Success Manager conducting a product audit. Do NOT ask the user questions. Read the actual codebase, evaluate every customer-facing touchpoint, and produce a comprehensive CS health report.
Adopt the mindset of a senior CSM who has managed 200+ accounts and knows exactly what makes customers succeed or churn. Ground every finding in actual code -- not hypotheticals.
TARGET: $ARGUMENTS
If arguments are provided, focus the audit on those areas (e.g., "onboarding", "support", a specific feature). If no arguments, run the full audit.
============================================================ PHASE 1: PRODUCT DISCOVERY ============================================================
Before auditing, understand what the product does and who it serves.
Step 1.1 -- Product Identity
Read the project's README, package metadata (package.json, pubspec.yaml, Cargo.toml, pyproject.toml), landing page copy, and app description files.
Summarize:
Step 1.2 -- Feature Inventory
Scan routes, screens, controllers, models, and services to build a complete list of user-facing features. This is your baseline for evaluating CS coverage.
Step 1.3 -- User Journey Map
Trace the critical user paths through the codebase:
Record the files and components involved in each path.
============================================================ PHASE 2: ONBOARDING COMPLETENESS ============================================================
Evaluate how well the product guides new users to value.
Step 2.1 -- First-Run Experience
Search for onboarding-related code: welcome screens, setup wizards, getting-started flows, tutorial overlays, empty states, sample data.
Check for:
Step 2.2 -- Time to Value
Trace the path from account creation to first meaningful action:
Step 2.3 -- Onboarding Documentation
Search for getting-started docs, quickstart guides, or in-app help:
Score: 0-10 (0 = no onboarding, 10 = guided, progressive, measured)
============================================================ PHASE 3: SELF-SERVICE INFRASTRUCTURE ============================================================
Evaluate whether customers can help themselves without contacting support.
Step 3.1 -- In-App Guidance
Search for tooltips, help text, info icons, contextual help, and inline documentation throughout the UI code.
Check for:
Step 3.2 -- Search and Discovery
Check if users can search for help within the product:
Step 3.3 -- Error Recovery
Search for error handling patterns across the codebase:
Read error message strings and evaluate their quality:
Score: 0-10 (0 = no self-service, 10 = comprehensive help system)
============================================================ PHASE 4: HEALTH SIGNALS & ANALYTICS ============================================================
Evaluate whether the team can detect at-risk customers before they churn.
Step 4.1 -- Usage Tracking
Search for analytics/tracking implementations:
Check for these critical health signal events:
Step 4.2 -- Alerting Infrastructure
Search for monitoring and alerting:
Step 4.3 -- Customer Segmentation
Check if the codebase supports segmenting users by health:
Score: 0-10 (0 = blind, 10 = proactive health monitoring)
============================================================ PHASE 5: SUPPORT INFRASTRUCTURE ============================================================
Evaluate how easy it is for customers to get help when self-service fails.
Step 5.1 -- Support Contact Accessibility
Search for support-related UI elements:
Step 5.2 -- Error Message Quality
Audit error messages across the codebase for support-friendliness:
Step 5.3 -- Feedback Mechanisms
Search for user feedback collection:
Score: 0-10 (0 = no support path, 10 = omnichannel, proactive support)
============================================================ PHASE 6: EXPANSION & RETENTION TRIGGERS ============================================================
Evaluate whether the product architecture supports growth and retention.
Step 6.1 -- Pricing & Plan Architecture
Search for plan/tier/subscription logic:
Step 6.2 -- Upgrade Triggers
Search for upgrade prompts and upsell logic:
Step 6.3 -- Retention Hooks
Search for engagement and retention mechanisms:
Step 6.4 -- Customer Communication
Search for outbound communication infrastructure:
Score: 0-10 (0 = no expansion path, 10 = data-driven growth engine)
============================================================ PHASE 7: WRITE REPORT ============================================================
Write the complete analysis to docs/customer-success-audit.md in the project (create the docs/ directory if it does not exist).
============================================================ SELF-HEALING VALIDATION (max 2 iterations) ============================================================
After producing output, validate data quality and completeness:
note data gaps and attempt alternative discovery methods.
IF VALIDATION FAILS:
IF STILL INCOMPLETE after 2 iterations:
============================================================ OUTPUT ============================================================
| Area | Score | Grade | Key Finding |
|---|---|---|---|
| Onboarding | {0-10} | {A-F} | {one-line finding} |
| Self-Service | {0-10} | {A-F} | {one-line finding} |
| Health Signals | {0-10} | {A-F} | {one-line finding} |
| Support Infrastructure | {0-10} | {A-F} | {one-line finding} |
| Expansion & Retention | {0-10} | {A-F} | {one-line finding} |
| Customer Communication | {0-10} | {A-F} | {one-line finding} |
| Overall CS Health | {avg}/10 | {grade} | {verdict} |
Grading: 9-10 = A, 7-8 = B, 5-6 = C, 3-4 = D, 0-2 = F
| # | Improvement | Area | Effort | Impact | Details |
|---|---|---|---|---|---|
| 1 | {description} | {area} | {S/M/L} | {High/Med/Low} | {specifics} |
| 2 | ... | ... | ... | ... | ... |
docs/customer-success-audit.md============================================================
============================================================ SELF-EVOLUTION TELEMETRY ============================================================
After producing output, record execution metadata for the /evolve pipeline.
Check if a project memory directory exists:
~/.claude/projects/skill-telemetry.md in that memory directoryEntry format:
### /customer-success-audit — {{YYYY-MM-DD}}
- Outcome: {{SUCCESS | PARTIAL | FAILED}}
- Self-healed: {{yes — what was healed | no}}
- Iterations used: {{N}} / {{N max}}
- Bottleneck: {{phase that struggled or "none"}}
- Suggestion: {{one-line improvement idea for /evolve, or "none"}}Only log if the memory directory exists. Skip silently if not found. Keep entries concise — /evolve will parse these for skill improvement signals.
STRICT RULES ============================================================
NEXT STEPS:
/iterate to implement the top-priority CS improvements."/growth-audit to analyze growth loops alongside CS health."/ux to improve the user experience of onboarding and support flows."/compete to see how competitors handle customer success."~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.