Find the best business email for this LinkedIn profile and tell me how fresh the result is.
example promptlibrary / resolution / resolve-one-linkedin-profile
Resolve one
LinkedIn profile.
Start with a LinkedIn URL. Return the best available business email together with freshness, validation risk, match confidence, provenance, and the exact token charge.
Draft Starter FetchSet MCP Claude
Resolved recordReady
Avery Lin avery@northstar.dev
98.4% confidence · 8 days fresh · 0 tokens
Example prompts
Ask for the outcome.
The skill handles the lookup sequence and returns the evidence your agent needs for the next decision.
Resolve this LinkedIn URL, but stop if the result is risky or the expected charge is above my limit.
example promptAbout this skill
What this skill does
This skill turns one LinkedIn profile URL into a business identity record that is useful without hiding uncertainty.
It validates the input, checks the expected token exposure, calls the narrowest FetchSet resolution tool, and returns the best available business email with the fields needed to decide whether the record is safe to use. Cache hits, external lookups, misses, and suppressed results remain explicit.
Resolution chainSKILL.md preview
LinkedIn profile URL → validate URL + normalize profile path → fetchset.account_get (confirm limits) → fetchset.linkedin_to_best_email (resolve best business email) → inspect freshness + risk + confidence → apply caller policy (use / review / stop) → return record + provenance + exact tokens charged
Inputs
Give the agent its boundaries.
The public profile URL that anchors the identity resolution request.
The maximum acceptable age of the last validation date, such as 30 days.
Whether catch-all, unknown, personal, or low-confidence results may be returned.
The maximum charge the agent may approve before it must stop or ask.
How it works
A short, inspectable path.
- Normalizes the LinkedIn URL and rejects company pages, search URLs, and malformed profile paths.
- Reads the FetchSet account state so the agent understands available tools, rate limits, and token balance before resolution.
- Runs the LinkedIn-to-best-email resolution tool using the caller’s allowed email types and cost boundary.
- Evaluates freshness, validation risk, and identity confidence as separate signals instead of collapsing them into one score.
- Returns a usable record, a review recommendation, or a structured miss, with the exact tokens charged in every case.
Output
The record carries its evidence.
| Name | Best email | Risk | Validated | Confidence | Source | Tokens |
|---|---|---|---|---|---|---|
| Avery Lin | avery@northstar.dev | Valid | 8 days ago | 98.4% | Fresh cache | 0 |
1Profile checked
1Usable record
0Tokens charged
Tips
Keep the policy visible.
- Define freshness and risk rules in the prompt instead of asking the agent to infer what “good” means.
- Treat identity confidence and email validation as separate decisions; a correct person match does not guarantee a deliverable address.
- Preserve structured misses. A suppressed or not-found result should stay empty and should show a zero charge.
- Use a batch workflow for lists. This skill intentionally resolves one profile so the decision trail stays easy to inspect.
Add to your agent
Install the path. Keep control of the call.
This page is a product draft. The downloadable package will be published after the skill schema and FetchSet tool names are stable.
Review installation paths01Connect the FetchSet MCP server
02Add the skill from the FetchSet plugin or skills.sh
03Start a new agent session and use an example prompt
Related skills
Continue from a resolved identity.
002 / Resolution
Choose the best available email
Apply email-type, freshness, and risk policy to more than one candidate.
Browse resolution 005 / Data quality
Audit email freshness
Find stale records and produce the smallest queue that needs revalidation.
Browse data quality 011 / Workflow
Enrich a LinkedIn CSV
Move from one profile to bounded batches without losing misses or charge evidence.
Browse workflows