hai-agents

A guide for delegating self-contained browser and research tasks to H Company’s remote computer-use agents. The agents browse and act on their own infrastructure rather than on the user’s computer.

In plain words
What is it for?
Use it to research across websites, extract information, navigate multi-step pages, and fill in or submit forms when the task is fully described.
Why use it?
It helps offload open-ended web work that requires navigation, multiple pages, form submission, or browser interaction.

Skill for Claude CodeCodex

Install

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.

agentmods
npx agentmods add skills/hcompai/hai-agents-python/hai-agents
Any agent
npx skills add hcompai/hai-agents-python --skill hai-agents
Clone the repo
git clone --depth 1 https://github.com/hcompai/hai-agents-python

Made for: Claude Code, Codex.

Per session 102 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 889 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What 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.

ModelPer sessionOnce invoked
Fable 5 $0.00102 $0.00889
Opus 5 $0.00051 $0.00445
Sonnet 5 $0.00020 $0.00178
Haiku 4.5 $0.00010 $0.00089

Measured 2d ago against content hash 3391d279a9aa, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

hai-agents scanned grade A with 1 finding 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

User: *"is the H Company Agent API quickstart still showing the curl example"*
src/hai_agents_cli/host_skills/hai-agents/SKILL.md · 46 lines

How it starts

The opening of the file, as written. The whole thing — 46 lines — stays where its author put it; the contents beside it link to each section on GitHub.

hai-agents

The hai-agents MCP server runs H Company's autonomous agents on H Company's infrastructure and streams back a single answer. You hand an agent a self-contained task; it opens its own browse-and-act loop in a remote browser and works until it has an answer, hits its step or time budget, or is cancelled. Nothing runs on the user's machine.

The agent is blind to this conversation. It sees only the task string, so fold in the context it needs: which site or source, what "done" looks like, the shape of the answer you want back. Keep the user's own action verbs and any literal text (search queries, message bodies) verbatim, since the agent grounds well on natural imperatives. Paraphrase by adding context, not by rewording.

Use it for open-ended work on the live web: finding and extracting information across pages, filling in and submitting forms, multi-step navigation behind logins the agent can perform, and research that needs a real browser. It is the wrong tool when the answer is already in your knowledge or a single fetch, when the job is local file, shell, or code work, or when the user wants something written or explained in the conversation rather than done on the web.

Tools

  • list_agents() — agents the caller can run (their org's plus the public h/ ones). Pick the agent name from here.
  • run_agent(task, agent, max_steps?, max_time_s?, idempotency_key?) — start a run. Returns either the final answer or a session handle { session_id, status, answer, done }.
  • wait_for_session(session_id, wait=True) — long-poll a running session for its answer; wait=False returns the current snapshot without blocking.
  • send_message(session_id, message) — steer a running session with a follow-up.
  • cancel_session(session_id) — stop a run you no longer need.
  • share_session(session_id) — get a public read-only URL for the run.

Long-running tasks

MCP clients cap a single tool call at roughly a minute, but agent runs often take longer, so the server long-polls instead of blocking the whole time:

Read the full file on GitHub · 46 lines

Changes

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.

  1. 2d ago First seen · 46 lines · 102 tokens per session scan A 3391d279a9aa

Subscribe to this mod's changes

hai-agents is a skill published in the GitHub repository hcompai/hai-agents-python (33 stars, last pushed 5d ago), licensed MIT. It adds 102 tokens to every session and 889 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

general-video

Author or edit a custom HyperFrames composition when no specialized workflow fits, or when BRIEF.md sets flow: companion. Use for longer or multi-scene pieces, brand and sizzle reels, montages, static loops, static title cards, footage remixes, and freeform builds. Use motion-graphics instead for a short unnarrated…

heygen-com/hyperframes · 92 tokens

use-agent-browser-for-airi

Test AIRI display-model imports with agent-browser across stage-tamagotchi Electron, stage-web, and stage-pocket mobile web layouts. Use when uploading and verifying contributor-supplied Live2D ZIP, VRM, or MMD ZIP/PMX/PMD files through AIRI's model selector, including onboarding bypass, format-specific import…

moeru-ai/airi · 87 tokens

opencli-sitemap-author

Use when creating or maintaining OpenCLI site sitemaps: agent-facing navigation, page-state, action, workflow, API-reference, pitfall, and fallback knowledge for a website. Use after browser exploration discovers durable site context, when a sitemap is stale, or when promoting local site knowledge into the repo.

jackwener/OpenCLI · 67 tokens

debug-optimize-lcp

Guides debugging and optimizing Largest Contentful Paint (LCP) using Chrome DevTools MCP tools. Use this skill whenever the user asks about LCP performance, slow page loads, Core Web Vitals optimization, or wants to understand why their page's main content takes too long to appear. Also use when the user mentions…

ChromeDevTools/chrome-devtools-mcp · 99 tokens

pinchtab-mcp

Use this skill when a task requires browser automation through PinchTab's MCP server connected to a remote browser instance. Covers navigation, element interaction, data extraction, form filling, multi-step flows, and session management via MCP tools.

pinchtab/pinchtab · 52 tokens

mintlify-preview

Run the public Mintlify product docs site locally for live preview. Use when previewing or iterating on docs under docs/ (.mdx/.md pages, docs.json nav), or when the user says "start the docs", "run mintlify", "preview the docs site".

latitude-dev/latitude-llm · 69 tokens