Stats
Actions
Tags
From fiber
Resolve GitHub handles or profile URLs into LinkedIn profiles and full contact enrichment using Fiber AI. Use this when the user says "I have GitHub handles", "enrich these engineers by GitHub", "find LinkedIn for these GitHub users", "developer enrichment", "OSS contributors list", "dev candidate enrichment", or otherwise starts from GitHub identifiers and wants LinkedIn or contact data.
How this skill is triggered — by the user, by Claude, or both
Slash command
/fiber:enrich-github-handles <one or more GitHub handles or github.com/USER URLs><one or more GitHub handles or github.com/USER URLs>The summary Claude sees in its skill listing — used to decide when to auto-load this skill
Developer-sourcing skill. Start from GitHub handles or profile URLs, resolve to LinkedIn, then optionally layer on email / phone / live profile enrichment. All GitHub resolution is async.
Developer-sourcing skill. Start from GitHub handles or profile URLs, resolve to LinkedIn, then optionally layer on email / phone / live profile enrichment. All GitHub resolution is async.
/fiber:enrich-linkedin-csv/fiber:find-and-enrich-by-role/fiber:build-recruiting-audience/fiber:sdk-ts or /fiber:sdk-pyRoute by what the user actually needs back. Both GitHub operations are async - expect to poll.
githubToLinkedInTrigger, save the returned task id, then poll githubToLinkedInPolling no more often than every 5 seconds until done. This is the cheapest path and returns just the matched LinkedIn URL.githubLookupTrigger, save the returned task id, then poll githubLookupPoll no more often than every 5 seconds until done. This returns a resolved person record (LinkedIn, name, title, company, location).syncQuickContactReveal per row, or startBatchContactDetails + pollBatchContactDetails for batches of 10-2000.profileLiveEnrich.For batches of 20+ handles, loop triggers sequentially but batch your polling - poll once per 15 seconds across all outstanding task ids.
Always call get_endpoint_details_full("<operationId>") on the Core MCP before constructing bodies. GitHub handle normalization (case, leading @, full URL vs bare handle) matters - check the schema.
githubToLinkedInTrigger / githubToLinkedInPolling: charges credits only on a successful match. Polling is free. Before triggering: "Resolve these N GitHub handles to LinkedIn URLs? Charges apply only to matches."githubLookupTrigger / githubLookupPoll: charges per successful resolution (more than the LinkedIn-only path because it returns a full person record). Only use when the user explicitly needs more than the LinkedIn URL.syncQuickContactReveal / startBatchContactDetails: charge per reveal. Confirm scope before running - see /fiber:enrich-linkedin-csv for the exact gates.profileLiveEnrich: charges per call. Confirm on larger fan-outs.getOrgCredits before starting a large batch.Never silently escalate from path 1 to path 2 - if the user asked for "LinkedIn URLs for these handles", do not also pull contact reveals without asking.
PENDING for > 5 minutes on a single task: keep polling but tell the user and offer to cancel. Never re-trigger an in-flight task; you will double-charge on success.torvalds, @torvalds, https://github.com/torvalds): normalize before triggering. The API accepts full URLs; bare handles may need the https://github.com/ prefix.triggerExhaustiveContactEnrichment as the waterfall fallback.Set user expectations up front - GitHub-to-LinkedIn match rates are structurally bounded by what the engineer has exposed publicly:
For developer-sourcing workflows, expect to lose 30-50% of the input list on the GitHub-to-LinkedIn hop alone. Plan the contact-reveal budget off the resolved count, not the input count. If the user needs every handle resolved, no amount of retries will help - gracefully surface the misses and move on.
Accept: text/markdown to https://api.fiber.ai/openapi.json to get the agent-friendly index at the same URL (Stripe pattern).This skill is primarily MCP-driven because polling logic fits naturally into an agent loop. When embedding into a product, both trigger/poll pairs follow the same shape as other async Fiber operations. See /fiber:sdk-ts and /fiber:sdk-py for client setup, and reuse the polling cadence rules above - never poll faster than every 5 seconds.
npx claudepluginhub fiber-ai/fiber-ai-plugin --plugin fiberGuides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.