Self-Improving Sales
v1.1.0Captures pipeline leaks, objection patterns, pricing errors, forecast misses, competitor shifts, and deal velocity drops to enable continuous sales improveme...
⭐ 0· 49·0 current·0 all-time
byJosé I. O.@jose-compu
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
Capability signals
These labels describe what authority the skill may exercise. They are separate from suspicious or malicious moderation verdicts.
OpenClaw
Benign
high confidencePurpose & Capability
Name/description, documentation, scripts, and hook handlers consistently implement a sales-learning capture and promotion workflow (create .learnings/, log issues, promote to skills). Required resources and actions align with the stated goal; no unrelated cloud credentials, binaries, or external download/install steps are required by default.
Instruction Scope
SKILL.md and included hooks/scripts stay within the sales-improvement scope (reminders, scaffolding, lightweight detectors). The OpenClaw hook performs prompt injection by adding a virtual reminder file at agent bootstrap (this is intentional for the skill). The error-detector reads CLAUDE_TOOL_OUTPUT to flag keywords — this can surface sensitive content if that environment variable contains raw tool outputs. SKILL.md does explicitly advise anonymization and not logging PII.
Install Mechanism
No automated remote install or archive downloads are specified. The package is instruction-only for installation; included scripts and hook handlers are provided in the bundle. A manual git clone example references a GitHub repo but that is optional; there are no extract-from-URL or shortener downloads.
Credentials
Metadata declares no required environment variables, but scripts (error-detector.sh) rely on CLAUDE_TOOL_OUTPUT at runtime. While CLAUDE_TOOL_OUTPUT is a plausible runtime hook variable, it is not listed in requires.env. The hook scripts run with the agent's permissions and will see any tool output provided to that env var; the skill advises not to log PII but cannot enforce it. This is proportionate to the detector feature but worth noting before enabling PostToolUse hooks.
Persistence & Privilege
The skill does not request always:true or permanent elevated privileges. Hook installation is explicitly opt-in (copy + enable). When enabled, the hook injects a reminder into bootstrap context (prompt injection by design) which broadens the skill's runtime presence inside agent sessions — acceptable for the feature but something to enable intentionally.
Assessment
This skill is internally consistent for capturing and promoting sales learnings, but check a few things before enabling hooks: (1) Review the scripts (activator.sh, error-detector.sh, extract-skill.sh) yourself — they are small and included; (2) Prefer the minimal setup (UserPromptSubmit/activator) and avoid enabling PostToolUse unless you understand and accept that the error-detector will read CLAUDE_TOOL_OUTPUT (which may contain sensitive command/tool output); (3) Keep the hooks opt-in and enable them only in trusted environments; (4) If you enable PostToolUse, restrict matchers so detectors only run for CRM/CRM-export contexts to reduce accidental exposure; (5) If you copy from the remote git URL, verify the repository content and origin first. If you want, test in a throwaway workspace to observe behavior before deploying to production.Like a lobster shell, security has layers — review code before you run it.
latestvk97a7tv4wbdwr3a3ajhnwqv3r184t9xb
License
MIT-0
Free to use, modify, and redistribute. No attribution required.