# 5dive skills
Skills published by [5dive](https://5dive.ai?utm_source=github&utm_medium=referral&utm_campaign=skills-readme) for use with
[skills.sh](https://skills.sh/) and any agent harness that loads
`SKILL.md` bundles (Claude Code, openclaw, hermes, …).
| Skill | Purpose |
| ---------------------- | ------------------------------------------------------------- |
| [`5dive-cli`](https://github.com/5dive-ai/skills/blob/HEAD/5dive-cli) | Drive the local `5dive` CLI on a 5dive runtime VM. Lets an agent spawn, inspect, send to, and tear down sibling agents on the same host. |
| [`diagnose`](https://github.com/5dive-ai/skills/blob/HEAD/diagnose) | Self-diagnosis for 5dive agents. Runs auth, service, disk, memory, and skill-integrity checks on the VM and returns a root-cause report — so the agent can fix problems itself instead of asking the user. |
| [`compile-knowledge`](https://github.com/5dive-ai/skills/blob/HEAD/compile-knowledge) | Compile durable knowledge into interlinked-markdown stores the "karpathy method" way — atomic files, `[[links]]`, a maintained index — so the agent gets smarter over time instead of relearning. Memory-first; shared wiki only for teams. |
| [`openagent`](https://github.com/5dive-ai/skills/blob/HEAD/openagent) | Author your own OpenAgent persona and mint your shareable holo trading card. Self-service: write a `<id>.persona.yaml`, validate it, check your rarity tier (Common→Mythical) + completeness, render the PNG card, and PR into the character-packs registry. |
| [`copywriting`](https://github.com/5dive-ai/skills/blob/HEAD/copywriting) | Write, rewrite, and improve conversion-focused marketing copy for any page — homepage, landing, pricing, feature, product. Headlines, CTAs, value props, taglines, hero sections. Adapted from Corey Haines' marketingskills (MIT). |
| [`no-ai-slop`](https://github.com/5dive-ai/skills/blob/HEAD/no-ai-slop) | Edit drafts into sharper, more human writing while preserving the writer's voice, or detect AI-slop patterns without rewriting. Use to make a draft clearer, more direct, and less AI-sounding, or to check whether writing reads as AI. MIT, courtesy of Peter Yang. |
| [`ad-creative`](https://github.com/5dive-ai/skills/blob/HEAD/ad-creative) | Generate and iterate ad creative at scale — headlines, descriptions, and primary text for paid platforms — and refine based on real performance data. Adapted from Corey Haines' marketingskills (MIT). |
| [`find-loops`](https://github.com/5dive-ai/skills/blob/HEAD/find-loops) | Search the [agenticloops.dev](https://agenticloops.dev?utm_source=github&utm_medium=referral&utm_campaign=skills-readme) directory for an existing agentic loop (a recurring, scheduled agent) and install it. The loop-level analogue of `find-skills`. |
| [`loops`](https://github.com/5dive-ai/skills/blob/HEAD/loops) | Author or modify a portable [`LOOP.md`](https://agenticloops.dev?utm_source=github&utm_medium=referral&utm_campaign=skills-readme) — a trigger + skills + a prompt in one file any harness can install and run on a schedule. The loop-level analogue of `skill-creator`; to install a loop that already exists use `find-loops`. |
| [`playwright-e2e`](https://github.com/5dive-ai/skills/blob/HEAD/playwright-e2e) | End-to-end testing and click-verification of web apps with Playwright. Install, write specs, drive authenticated pages, screenshot, and run in CI. Use to prove a web change actually works in a real browser before shipping. |
| [`code-review`](https://github.com/5dive-ai/skills/blob/HEAD/code-review) | Review a diff or pull request for correctness first, quality second. Logic bugs, edge cases, error handling, races, missing tests, API misuse, performance, readability, with a severity model and clear approve vs request-changes verdict. |
| [`verify`](https://github.com/5dive-ai/skills/blob/HEAD/verify) | Grade a delivered claim against the artifact instead of against the report of it. Split a claim into checkable assertions, grade each against a named instrument, and emit a three-state verdict — pass, fail, or not-reached — with what you did *not* check stated explicitly. |
| [`deep-research`](https://github.com/5dive-ai/skills/blob/HEAD/deep-research) | Run structured, multi-source research that produces a trustworthy, cited synthesized answer rather than a pile of links. Scope the question, fan out across sources, judge source quality, adversarially verify each key claim, separate fact from inference, and report with explicit confidence. |
| [`zero-human-dev-team`](https://github.com/5dive-ai/skills/blob/HEAD/zero-human-dev-team) | Every rule 5dive's AI coding team works by: tasks, code, tests, review, releases, backups, secrets, and how agents talk to each other and their human. |
| [`caveman-marketing`](https://github.com/5dive-ai/skills/blob/HEAD/caveman-marketing) | Cut an ad, landing page or signup flow down until a stranger gets it in 3 seconds: one picture, at most 15 words, one button. Ships a word-and-button checker and real before/afters. |
## Quick start (any coding agent)
Install straight from the [skills.sh](https://skills.sh/) directory with the open
`skills` CLI — works in Claude Code, Codex, Cursor, and any harness that loads
`SKILL.md`:
```bash
# mint your own OpenAgent persona + shareable holo card
npx skills add 5dive-ai/skills --skill openagent
# find, install, and run agentic loops (recurring agents)
npx skills add 5dive-ai/skills --skill loops
# every rule our AI coding team works by
npx skills add 5dive-ai/skills --skill zero-human-dev-team
# make any ad, page or signup flow pass a 3-second glance
npx skills add 5dive-ai/skills --skill caveman-marketing
# or pull the whole set
npx skills add 5dive-ai/skills --skill '*'
```
`openagent` and `loops` are the best starting points — they work in any agent,
no 5dive account required. [`loops`](https://github.com/5dive-ai/skills/blob/HEAD/loops) pairs with the open
[agenticloops.dev](https://agenticloops.dev?utm_source=github&utm_medium=referral&utm_campaign=skills-readme)
directory + `LOOP.md` spec.
## Install on a 5dive agent
From the dashboard: **Agents → Connect skills → 5dive-cli → Install**.
From the host:
```bash
sudo 5dive agent skill <agent-name> add \
--source=5dive-ai/skills \
--skill=5dive-cli
```
Or via the upstream `skills` CLI from any agent's home:
```bash
npx skills add https://github.com/5dive-ai/skills --skill 5dive-cli --agent claude-code --yes
```
## License
MIT.
## Description budget
A skill's `name` + `description` is loaded at **selection** time, in every
session, whether or not the skill fires. A command catalog there is paid on
every turn and mis-routes the selector, so every description in this repo is
capped at **two discriminating sentences** — the capability, the activation
condition, and at most one exclusion. Command names, examples and feature
catalogs live in the skill body or in `references/`.
```bash
./scripts/check-descriptions.sh # enforces the cap, prints total chars
```
`evals/routing.json` holds the behavioural routing cases that guard the
boundaries the cap is there to protect — an unrelated failed command must not
load `diagnose`, a normal local coding task must not load `5dive-cli`,
searching for a scheduled workflow selects `find-loops` while authoring a
`LOOP.md` selects `loops`.
