Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/agentled/mcp-server/linkedin-assistantnpx skills add Agentled/mcp-server --skill linkedin-assistantgit clone --depth 1 https://github.com/Agentled/mcp-serverWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00031 | $0.01627 |
| Opus 5 | $0.00015 | $0.00813 |
| Sonnet 5 | $0.00006 | $0.00325 |
| Haiku 4.5 | $0.00003 | $0.00163 |
Grade A, and why
linkedin-assistant scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 129 lines — stays where its author put it; the contents beside it link to each section on GitHub.
LinkedIn Assistant
Purpose
Use this skill to maintain a LinkedIn operating ledger in AgentLed and to prepare actions that a human or approved executor can run later. The skill is for tracking, classification, drafting, quota control, and approval preparation. It is not permission to send LinkedIn messages, connection requests, InMail, email, WhatsApp, provider campaigns, CRM writes, or public posts.
Source Of Truth
Resolve the target workspace, use case, and operating workflow before reading or changing LinkedIn state:
agentled --workspace <workspace> workflows get <workflow-id> --format json
agentled --workspace <workspace> use-cases get <use-case-key> --format json
Read every operating guide listed in useCaseContext.useCase.operatingGuides before making state changes.
Identify the workspace's lists for active people, follow-up actions, activity events, daily quota, archived contacts, and campaign-session deduplication from those guides. Do not assume personal workspace aliases, UUIDs, or list keys.
Timestamp Contract
Do not let timestamp fields drift into vague "Date" semantics.
- Row
createdAtandupdatedAtare AgentLed storage metadata. They mean when the KG row was inserted or changed. rowData.occurredAtis the timestamp used for activity chronology. It must describe the LinkedIn lifecycle event time when that time is visible or explicitly confirmed.- If the exact LinkedIn event time is not known and the event was only observed or flagged by the user,
occurredAtmay be the observed/flagged time, but the row must also includetimestampType: "observed_at"and a shorttimestampNote. - If only a calendar date is known, use local midnight for that date and include
timestampPrecision: "date"plus a note. - Person fields such as
connectionSentAt,connectedAt,messageSentAt,replyReceivedAt,meetingScheduledAt, andmeetingDoneAtare lifecycle timestamps. Leave them blank unless there is visible LinkedIn evidence or explicit user confirmation. lastSentAtis the latest outbound LinkedIn touch for the person, including connection requests, first messages, and follow-ups. StorelastSentType,lastSentAtType, and a shortlastSentAtNotewhen the timestamp is observed/imported rather than verified from LinkedIn.lastReplyAtis the latest inbound LinkedIn reply from the person. StorelastReplyAtTypeand a shortlastReplyAtNotewhen the exact LinkedIn receive timestamp is not visible.messageSentAtandreplyReceivedAtcan keep a specific lifecycle milestone, but reports and follow-up decisions should uselastSentAtandlastReplyAtonce those fields exist.lastActivityAtmay point to the latest verified or observed activity, but do not use it as a replacement forlastSentAtorlastReplyAt. AddlastActivityAtTypewhen it is not a verified LinkedIn event timestamp.capturedAtor rowcreatedAtcan be used to explain when AgentLed learned about the event.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 129 lines · 31 tokens per session scan A a60515804d8c
linkedin-assistant is a skill published in the GitHub repository Agentled/mcp-server (3 stars, last pushed 12d ago), licensed MIT. It adds 31 tokens to every session and 1,627 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
lost-deal-revival-agent
Drafts revival messages for closed-lost deals when a public company signal contradicts the original objection. Receives the signal-pair watcher's fire payload, fetches the verbatim Claap objection quote, and produces a 2-line draft that quotes it back. Default output is a HubSpot task for human review. Optional…
cold-email-strategist
Use this skill when writing or fixing cold email — first-touch and follow-up sequences, re-engagement, subject lines, personalization at scale, ATL/BTL messaging, copywriting frameworks, deliverability, domain warmup, SPF/DKIM/DMARC, and email infrastructure. Triggers on 'cold email', 'email sequence'…
personalization-6-buckets
Use this skill when researching a prospect for personalized outreach, choosing a personalization angle — self-authored content, engaged content, self-identified traits, background, or company-level triggers — or falling back to core-static relevance when no personal data exists.
atl-btl-messaging
Use this skill when crafting messages for different seniority levels, deciding message length and tone, or adapting outreach to VPs and C-Level vs Managers and ICs — the Above-the-Line (ATL) / Below-the-Line (BTL) messaging framework for targeting executives vs end-users.
email-metrics-benchmarks
Use this skill when analyzing cold email campaign performance, setting benchmarks, diagnosing low open or reply rates, or optimizing outreach metrics. Covers 2X levers, reply rates by outreach type, deliverability thresholds, sequence performance, and quick diagnostics.
linkedin-success-factors
Use this skill when reviewing a LinkedIn campaign before launch, auditing LinkedIn outreach performance, checking targeting, message length, lead magnets, scheduling, and human-like behavior, or benchmarking connection acceptance and reply rates.