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 commands/arcassystems/founderos/linkedin-scangit clone --depth 1 https://github.com/ARCASSystems/FounderOSWhat 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 | $0.00116 | $0.00768 |
| Opus 5 | $0.00058 | $0.00384 |
| Sonnet 5 | $0.00023 | $0.00154 |
| Haiku 4.5 | $0.00012 | $0.00077 |
Grade A, and why
linkedin-scan 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 3d 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 — 38 lines — stays where its author put it; the contents beside it link to each section on GitHub.
LinkedIn network scan
One-command entry point for the linkedin-network-scan skill. Point it at your LinkedIn data export and get back a ranked worklist plus an interactive page, scored locally against an ICP you control. The raw CSVs never enter the conversation - a deterministic, LLM-free script collapses them to a compact digest first.
Argument: $ARGUMENTS - the path to the export ZIP or unzipped export folder, with an optional --icp <file>. If no path is given, ask for one.
Procedure (in order)
-
Read the skill at
skills/linkedin-network-scan/SKILL.mdand follow it end to end. The skill owns the flow: the export gate, the ICP choice, the local run, and reading only the compact digest. -
The whole run is one command the skill issues:
python skills/linkedin-network-scan/scan.py <export> <output-folder> --icp <icp.yaml><output-folder>MUST be outside any git repo - it holds real names and profile URLs. The user's home directory or Desktop is fine.--icpis optional. Two example ICPs ship with the skill (sales / partnership and career / talent). Omit it for the permissive default and say so.
-
After the run, read ONLY
network-scan.md(the compact ranked digest). Do not read the CSV, the JSON, or the raw export - that is the token waste this skill exists to avoid. Then summarise: how many connections were read, how many qualified, how many replied before, and how many invitations are pending. Point the user atnetwork-scan.html(anonymised demo, safe to record and share) andnetwork-scan-full.html(real names and links, local only). -
If the run prints a freshness warning (export older than ~30 days), relay it and the re-pull steps before treating the worklist as current.
Rules
- This command is a thin trigger. All logic lives in the skill. Do not duplicate the steps here.
- The scoring engine is deterministic and free-plan only - no scrapers, no paid tools, no API calls. Only optional narrative or synthesis afterward uses the session's own Claude.
- Never analyse an export the user did not personally download. Refuse on consent grounds.
- Output files hold real names and URLs. Tell the user to keep the output folder local and never commit it.
- If the skill file is missing, reply:
LinkedIn scan skill not found at skills/linkedin-network-scan/SKILL.md. Restarting Claude Code fixes this most of the time, because it reloads what is installed. If it happens again after a restart, say "update Founder OS".and stop. - No em dashes or en dashes. Hyphens only with spaces.
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.
- 3d ago First seen · 38 lines · 116 tokens per session scan A de89e236cfac
linkedin-scan is a command published in the GitHub repository ARCASSystems/FounderOS (5 stars, last pushed 19d ago), licensed MIT. It adds 116 tokens to every session and 768 once invoked, about $0.0006 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 commands, from other repositories
wiki-init
Initialize a new LLM Wiki in the current directory. Creates the full directory structure, config, and template files.
check-invariants
Runs only the tree CRDT invariant test battery in outl-core. Faster than /check, focused on what can break sync.
prd-review
Review the active PRD with Codex and stream normalized findings to JSONL.
claude-cmd-status
A Claude Code slash command. OpenCode does not read Claude commands.
ingest-l1
L1 analysis loop for the abapwiki knowledge base: for each batch it launches the abap-analyzer sub-agent in parallel, then the adversarial judge abap-deepcheck (separate session), applies only the analyses that pass the fail-closed gate, and commits. Resumes exactly after an interruption. Use this skill to document…
web-search
Search fetched web-source Markdown by regex and pull surrounding context for the best hits. Optionally restrict to one source alias.