SpeciLiam
- 28 skills
- 0 followers
- 4 hours ago last updated
- ▌ Tandem · speciliam bundleDrive a coding task with BOTH Claude and Codex collaborating for better-planned, better-executed output. Invoke as $tandem or /tandem from either the Codex CLI/UI or Claude, in any repository. Whichever model you launch from becomes the driver; it pulls in the other model as its peer to (1) independently critique the plan, (2) optionally split or delegate implementation slices, and (3) review the final diff. Use for non-trivial or long-winded changes where a single model tends to drift, miss edge cases, or under-plan — e.g. 'I want this change in my fantasy-wizard repo and I want both models on it.'
- ▌ Job Intake · speciliam bundleRun scheduled LinkedIn and Greenhouse job intake, dedupe fresh software engineering roles Liam is realistically qualified for against Liam Van's tracker, rank them with NYC/SF preferences, and queue the best roles for tailoring/apply follow-through.
- ▌ Conduct Drain · speciliamClaude-conducted, Codex-executed application drain. Claude is the conductor — it builds and audits the queue, drains it in small bounded batches of Codex CLI workers (reusing finish-app-script's run_queue.py), runs a reasoning checkpoint between every batch to catch tracker corruption and systemic browser failures, gates the shared git push, and reconciles the tracker at the end. Use when the user says `$conduct-drain`, "conduct the drain", "drain my applications with review", or wants the long application drain run with Claude judgment at the seams instead of an unattended Python loop.
- ▌ Resume Tailor · speciliam bundleTailor a one-page resume from a generic Overleaf LaTeX resume using a job description URL or pasted posting. Use when the user wants to reorder skills, trim irrelevant content, strengthen bullets, remove weaker roles, and create a company-specific resume folder while keeping the output truthful and concise.
- ▌ Conduct Apply All · speciliamClaude-conducted, Codex-executed LinkedIn apply-all drain. Claude (Opus 4.8) is the conductor — it sets up the search/queue state, then drains a LinkedIn software-engineer search one posting at a time by launching a fresh single-application Codex worker (reusing linkedin-apply-all's run_queue.py), runs a reasoning checkpoint after every single application to catch tracker corruption, bad dedupe, wrong-resume drift, and systemic browser failures, gates the shared git push, and reconciles the tracker at the end. Each Codex worker tailors the resume for its posting first (resume-tailor), then applies, then exits. Use when the user says `$conduct-apply-all`, "conduct the apply-all", "drain LinkedIn with review", or wants the LinkedIn search drained with Claude judgment between every application instead of an unattended Python loop.
- ▌ Finish App Script · speciliam bundleDrain Liam's tailored-resume application queue from a phone-friendly command. Build `/tmp/fa_script_run_state.json`, then launch one fresh Codex or Claude worker per application so each worker attempts one form, submits and closes the tab when high-confidence, leaves a prepared handoff tab on true blockers, updates tracker/cache, and exits. Use when the user says `$finish-app-script`, "run finish app script", "drain my application queue", or wants applications completed without context overflow.
- ▌ Linkedin Outreach · speciliam bundleBuild, draft, and record LinkedIn outreach for any contact lane (recruiter, engineer, alumni, hiring manager, founder, peer) tied to Liam Van's tracked applications. Uses standing rules and per-lane signals to draft notes, queue approved rows, and update the tracker after each send.
- ▌ Linkedin Apply All · speciliam bundleDrain Liam Van's LinkedIn software engineer search results one by one from Chrome with exactly one active worker/browser actor, using a provided LinkedIn jobs search URL or the salary-filtered SWE search, skipping duplicate/already-applied postings, tailoring or reusing resumes in the same worker, submitting safe applications, recording precise blockers, and updating the markdown tracker/cache.
- ▌ Company Prospecting · speciliam bundleBuild and maintain a separate company-to-people outreach prospect tracker for job applications. Use when you want the top 3 recruiter, alumni, engineer, or employee contacts per company in markdown, plus Apollo email lookup status that stays separate from application-trackers/applications.md.
- ▌ Finish Applications · speciliam bundleComplete Liam Van's tracked job applications that have tailored resumes but are not yet applied. Use when the user wants one Codex agent to work through `application-trackers/applications.md` rows with `Applied` blank/false and `Status` like `Resume Tailored`, handle live Chrome application flows directly, submit where possible, ask only for blocking form answers or consent-sensitive choices, update the markdown tracker, refresh the recruiting dashboard cache, and perform a light file-based handoff only when context is overloaded.
- ▌ Greenhouse Sourcing · speciliam bundleSource high-volume SWE I, SWE II, early-career, new-grad, founding engineer, generalist, backend, full-stack, and forward-deployed roles from MyGreenhouse or Greenhouse job boards, dedupe them against Liam Van's CodexSkills application tracker, rank fit, queue resume tailoring, and apply through Greenhouse autofill where possible.
- ▌ Recruiting Pipeline · speciliam bundleOrchestrate Liam Van's full recruiting workflow across resume tailoring, application tracking, Gmail status refreshes, LinkedIn recruiter and engineer outreach, company prospecting, Apollo email lookup, and the Vercel visualizer, with optional routing to the separate Notion mirror skill only when requested.
- ▌ Bay Area Housing Hunt · speciliam bundleFind, normalize, score, and maintain Liam Van's Bay Area housing search around the HackerRank Santa Clara office. Use when Liam asks to search Bay Area housing, scrape or review Facebook Marketplace/Craigslist/apartment listings, compare SF neighborhoods or Peninsula/South Bay cities, compute commute and car scenarios, refresh daily top-5 power rankings by city/neighborhood, mark expired listings, or manage housing automation/tracker files.
- ▌ Linkedin Conduct Drain · speciliam bundleCodex-conducted LinkedIn early-career weekly drain with checkpoint review between one-stage workers. Use when Liam says `$linkedin-conduct-drain`, asks to conduct the LinkedIn early-career weekly flow, wants last-week Entry-level software-engineer LinkedIn postings discovered, deduped, tailored, and applied with Codex checkpoint judgment between stages, or wants the `linkedin-early-career-weekly` workflow to use the shared application-answer context while still updating the canonical tracker, manual handoff file, visualizer cache, and weekly durable state.
- ▌ Linkedin Full Pipeline · speciliam bundleRun Liam Van's all-in-one LinkedIn job pipeline from an early-career-focused LinkedIn software engineer search, including overnight-style drains: capture fresh preferred-location job postings, dedupe against the application tracker, tailor and verify a resume, add the role to the markdown database, verify recruiter LinkedIn outreach, send invites only until LinkedIn throttles connection requests, continue applications after outreach throttling, sort application tabs by confidence when possible, update tracker/cache state, and report precise blockers.
- ▌ Linkedin Outreach Batch · speciliam bundleThin compatibility entrypoint for preparing approved LinkedIn outreach batches through the generalized linkedin-outreach lane workflow.
- ▌ Notion Application Sync · speciliam bundleOptionally mirror Liam Van's application tracker data into Notion from the generated visualizer JSON cache. Use only when the user explicitly asks to sync, audit, configure, or schedule Notion matching; normal recruiting skills should not call this by default.
- ▌ Dual Agent Orchestration · speciliamCoordinate Codex and Claude Code as an executor/reviewer pair using local handoff files, Claude subagents, and shared recruiting workflow skills without letting both agents edit the same files at once.
- ▌ Gmail Application Refresh · speciliam bundleReview Gmail for application-related emails, infer status changes such as applied, online assessment, interview, rejection, or offer, and update the markdown application tracker without creating duplicate or low-confidence changes. Notion mirroring is handled separately by notion-application-sync.
- ▌ Linkedin Unattended Drain · speciliam bundleLeave-the-PC LinkedIn early-career drain — one supervised 60-second kickoff (Chrome check + one request_directory approval + watchdog start), then Claude (Fable 5 / Opus-class) drains discover → tailor → apply fully hands-off until saturation or caps. Never pauses for human gates — every manual blocker is parked to the handoff file and the run continues. A background watchdog keeps the Mac awake, guards RAM, reconciles the tracker/SQLite/visualizer deterministically, and sends macOS notifications on completion or systemic failure. Use when Liam says `$linkedin-unattended-drain`, "apply while I'm gone", "run it and I'll leave my PC", or wants the weekly drain to survive manual blockers without stopping.
- ▌ Linkedin Batch Drain Codex · speciliam bundleCodex-conducted LinkedIn batch drain for Liam's early-career SWE search. Use when Liam says $linkedin-batch-drain-codex or wants Codex chat to act as a persistent monitor that first discovers the requested target of fresh non-duplicate LinkedIn postings, normally 40 when Liam asks for a 40-role batch, then tailors verified resumes efficiently with no-browser workers, and finally applies sequentially in ATS-friction order while submitting high-confidence applications and parking only true blockers.
- ▌ Linkedin Early Career Weekly · speciliam bundleDurable LinkedIn early-career weekly application drain that avoids the older LinkedIn all-apply flow, starts from last-week entry-level software-engineer results, dedupes, spawns fresh Codex workers for discovery, resume tailoring, and one application at a time, and resumes from /tmp state through context degradation.
- ▌ Linkedin Easy Apply Nodriver · speciliam bundleUse this when Liam wants a nodriver-first LinkedIn sourcing pass from the signed-in Ben Chrome profile: find early-career software engineering jobs, mark LinkedIn Easy Apply postings for manual review, click through non-Easy-Apply external links, then switch actual ATS/application work to Liam's Chrome profile, tailor and upload resumes, fill every safe required field, avoid cover letters unless required, dedupe against the tracker, and stop as soon as a duplicate posting is encountered.
- ▌ Application Visualizer Refresh · speciliam bundleRefresh the Vercel-ready application tracker visualization website by parsing application-trackers/applications.md and outreach-prospects.md into application-visualizer/src/data/tracker-data.json.
- ▌ Linkedin Batch Drain Codex Ver2 · speciliam bundleCompaction-safe Codex-conducted LinkedIn batch drain for Liam's early-career SWE/FDE search. Use when Liam says $linkedin-batch-drain-codex-ver2, asks for the batch drain v2/ver2, or wants a large LinkedIn batch run with durable monitor checkpoints, one browser actor, child Chrome preflight for spawned application workers, and inline fallback when spawned workers cannot use the Chrome plugin or upload resumes.
- ▌ Linkedin Unattended Drain Codex · speciliam bundleCodex-driven leave-the-PC LinkedIn early-career drain — an interactive Codex (GPT 5.5) session is the conductor AND the browser actor (Codex Chrome plugin first, Computer Use fallback), with best-effort upload-origin warming at a supervised kickoff, park-don't-pause on every manual blocker, isolated /tmp state, and the shared path-parameterized watchdog for caffeinate/RAM/stall/reconcile/notifications. Use when Liam says `$linkedin-unattended-drain-codex`, wants the unattended drain run by Codex instead of Claude, or wants to compare the Codex upload path (absolute local paths, per-origin approvals) against the Claude variant's one-grant model.
- ▌ Linkedin Early Career Weekly Claude · speciliamClaude-conducted, Codex-executed version of the LinkedIn early-career weekly drain. Claude (Opus 4.8) is the parent/conductor — it builds the durable run state, then drains last-week Entry-level software-engineer results one stage at a time by launching a single fresh Codex worker per stage (reusing linkedin-early-career-weekly's run_stages.py with --max-stages 1), runs a reasoning checkpoint after every discover/tailor/apply stage to catch bad dedupe, wrong-resume drift, tracker corruption, and systemic browser failures, and reconciles the tracker at the end. Each Codex worker owns exactly one stage for one posting, then exits. Use when the user says `$linkedin-early-career-weekly-claude`, "conduct the early-career weekly with Codex workers", or wants the weekly drain run with Claude judgment between every stage instead of the blind run_monitored.py loop.
- ▌ Linkedin Early Career Weekly Claude Only · speciliam bundleClaude-only version of the LinkedIn early-career weekly drain — no Codex at all. Claude (Opus 4.8) is the conductor AND every worker is a fresh Claude subagent (Agent tool). Claude builds the durable run state into an isolated /tmp file, then drains last-week Entry-level software-engineer results one stage at a time by launching one fresh Claude subagent per discover/tailor/apply stage, runs a reasoning checkpoint after every stage, and reconciles the tracker at the end. Browser work uses Claude-in-Chrome / Computer Use (never the Codex Chrome plugin, never codex exec). Use when the user says `$linkedin-early-career-weekly-claude-only`, "run the weekly with only Claude workers", or wants the weekly drain done entirely by Claude with no Codex.