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 skills add open-octo/octo-agent --skill deep-researchgit clone --depth 1 https://github.com/open-octo/octo-agentWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/open-octo/octo-agent/deep-research)<a href="https://agentmods.dev/skills/open-octo/octo-agent/deep-research"><img src="https://agentmods.dev/badge/skills/open-octo/octo-agent/deep-research/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/open-octo/octo-agent/deep-research"><img src="https://agentmods.dev/badge/skills/open-octo/octo-agent/deep-research.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00158 | $0.01425 |
| Opus 5 | $0.00079 | $0.00713 |
| Sonnet 5 | $0.00032 | $0.00285 |
| Haiku 4.5 | $0.00016 | $0.00143 |
Grade A, and why
deep-research 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 10d 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 — 125 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Skill: deep-research
A harness for research you can trust: breadth first (many angles), then depth
(primary sources), then an adversarial pass that tries to break each claim
before it goes in the report. Built on octo's native tools — web_search,
web_fetch, sub_agent, and (for login-gated / JS-rendered / anti-bot sites)
the browser tool via the web-access skill. No external services.
The goal is a cited report where every non-obvious claim traces to a source you actually read — not a plausible-sounding summary of search snippets.
0. Scope before you search
Research is only as good as the question. Before any tool call, confirm you can state the deliverable: what question, what decision it informs, what time window, what region/market, what depth. If any of these is missing and would change the answer, ask 2-3 sharp clarifying questions first — don't guess a scope and burn a fan-out on the wrong one.
Then write down, in one line, what "done" looks like. That line is the acceptance criterion the final report is checked against.
1. Fan out — breadth
Decompose the question into 4-8 independent sub-questions, each attacking a different angle (definition, current state, competing views, data/numbers, history, criticisms, primary actors). Independence matters: overlapping sub-questions waste the fan-out.
Dispatch them in parallel. web_search / web_fetch are stateless, so this is
exactly the case sub-agents are for:
- One
sub_agentper sub-question. Prompt it goal-first, not step-first: describe what to find out, not "search for X" — an anti-bot source may needbrowseron the main site, and "search" would anchor the sub-agent toweb_search. - Tell each sub-agent to load the
web-accessskill and follow it, to return findings with source URLs, and to flag anything it couldn't verify. - Do NOT parallelize
browserwork — the browser session is single-page and process-shared; concurrent sub-agents fight over one page. Keep browser interaction to a single sequence; parallelize only the stateless search/fetch.
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.
- 10d ago First seen · 125 lines · 158 tokens per session scan A fab92cae827f
deep-research is a skill published in the GitHub repository open-octo/octo-agent (97 stars, last pushed yesterday), licensed MIT. It adds 158 tokens to every session and 1,425 once invoked, about $0.0008 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 2026-08-30.
Other skills, from other repositories
ha-skill-creator
Create, edit, improve, or audit Hope Agent skills. Use when the user wants to: (1) create a new skill from scratch, (2) edit or improve an existing skill, (3) review or clean up a SKILL.md file, (4) run evaluations to test skill effectiveness, (5) optimize skill descriptions for better trigger accuracy. Trigger…
ha-mac-control
Hope Agent native macOS desktop control — the standard maccontrol status / diagnostics / apps / dock / spaces / snapshot / visual / windows / menu / clipboard / dialog loop, target-first action rules, no-blind-coordinate policy, and recovery for stale AX/window/menu/dialog state. Load whenever using maccontrol, or…
ha-browser
Hope Agent browser automation — the standard status → tabs → snapshot → act loop, stale-ref recovery rules, and what to do when login / 2FA / captcha / camera-prompt / dialog blocks progress. Load this skill whenever you reach for the browser tool. Trigger on: user asks the agent to open / control / click / scrape /…
ha-logs
A read-only troubleshooting skill for querying Hope Agent’s local SQLite databases, which store logs, conversations, and background-job status.
ha-data-analytics
A local-first data-analysis and reporting skill for CSV and spreadsheet files. It produces decision-ready analyses and shareable offline reports while separating facts, calculations, interpretations, and recommendations.
ha-pet-import
Safely import, select, switch, or enable a compatible desktop pet in Hope Agent. Resolve packages from any origin, including local folders, zip archives, pet.json plus a sprite, PNG/WebP atlases, chat attachments, repository or cloud files, direct HTTPS artifact URLs, and download pages. Use whenever a user asks to…