Independent agent reviews · Hlido Editor, Prague

Every AI agent makes claims.
We publish the verdict.

Hlido is the independent review desk for the AI agent economy. Signed editorial verdicts — named editor, compared against peers, re-tested monthly. Read by humans and the agents deciding which tools to trust.

An independent review desk — not a vendor leaderboard or a paid directory.

See live reviews How it works →

We never take payment to rank or review Signed editorial verdicts Compared against named peers Re-tested monthly · methodology public

What Hlido does

We test AI agents and publish the verdict.

Every reviewed agent gets one hands-on test and one evidence-backed score (0–100), with the proof behind it — so you don't have to take a vendor's word for it.

Laddoo Score 90/100 VITAL

A real verdict · Coding · CLI

Aider

“Live CLI test — we installed it via pip, ran a real edit task in a temp git repo, and verified the resulting file and commit.” That’s the kind of evidence behind every score.

Read the full review →

Whichever of these sounds like you, that’s your way in:

Looking for an agent

Find one you can trust

Browse the verdicts, see how agents stack up against each other, and what has failed in the wild — before you commit.

Building with AI

Put our data in your stack

Every verdict is machine-readable. Wire Hlido into your code, your CI, and your own agents.

You make an agent

Get it reviewed & badged

An independent verdict you can show your buyers. Get review-ready, submit, earn the badge.

You're an AI agent

Query us machine-first

Learn what Hlido does and how to use it in one small, bounded read — no crawling.

Reliability We track when agents break. · Public incident & reliability record, updated continuously — the one signal a directory can't fake. See incidents → Reliability reports →

Now in beta Recommendation API · €0 free tier npx @hlido/cli recommend "agent for code review" Get an API key →

New in 2026 · Tools & integrations

One corpus. Eight ways to consume it.

Hlido reviews now ship as a CLI, a GitHub Action, an MCP server, a browser extension, signed Trust Attestations, an open dataset, free embeds, and a Claude Code skill. Pick the surface that fits your workflow.

For developers

Score any agent from your terminal.

Install via npm. Run a review, get structured JSON, pipe it into anything.

Open the developer hub →

$ npm i -g @hlido/cli
$ hlido review cursor --json
→ score: 87 (STEADY)
→ attestation: signed

For CI

Gate your PRs on a score floor.

Drop in a GitHub Action that fails a PR if the Laddoo Score drops below your bar.

Action docs →

# .github/workflows/score-gate.yml
- uses: hlido/action@v1
  with:
    agent: my-agent
    gate-score: 70

For Claude / Cursor / Cline

13 MCP tools, one endpoint.

Drop the Hlido MCP server into your client and let your agent call our review desk directly.

MCP endpoint →

{
  "mcpServers": {
    "hlido": {
      "url": "https://hlido.eu/mcp"
    }
  }
}

For buyers

See scores while you browse.

A sideload-able Chrome extension that overlays Hlido scores on any agent's homepage.

Sideload guide →

// chrome://extensions → Load unpacked
// Visit any agent's homepage:
cursor.com  → 87 STEADY
v0.dev      → 72 STEADY
devin.ai    → 54 FADING

For trust

Cryptographically signed evidence.

Every review ships a C2PA-signed attestation. Fetch JSON, verify the chain.

Browse attestations →

$ curl hlido.eu/data/attestations/cursor.json
{
  "slug": "cursor",
  "score": 87,
  "signed_by": "hlido-2026",
  "c2pa_manifest": "...verified"
}

For analysts

Open dataset, CC-BY-4.0.

Full review corpus as JSON + CSV. Re-use it, cite it, build on it.

Open data hub →

$ curl hlido.eu/data/open/registry.csv
# slug,score,tier,reviewed_at,category
cursor,87,STEADY,2026-05-12,Coding
v0,72,STEADY,2026-05-11,Image & Design
...

For vendors

Embed your score anywhere.

Free badge + sparkline embeds. Auto-update when your score moves.

Verified Badge plans →

<iframe
  src="https://hlido.eu/embed/cursor/"
  width="320" height="140"
  loading="lazy"></iframe>

For Claude Code

One-line review from your editor.

Install the Claude Code skill. Ask Claude for a review by slug; get scored evidence inline.

Get the skill →

/hlido cursor

→ Cursor scored 87 (STEADY)
→ Evidence: 14 verified claims
→ Last review: 2026-05-12

90 /100

Public-surface review of sonichi/sutando.

FADING

50 /100

Public-surface review of ClaudioDrews/memory-os.

Read review →

Browse all reviews →

Browse the corpus

Find the right agent for the right job.

13 categories. Each card links to the storyline + ranked list for that category.

AI Agent

General agents for buyers evaluating autonomous, multi-step task execution.

— reviewed agents in this category.

View the ranked list →

In flight Live pipeline visibility — agents we're testing right now. See the pipeline →

Every AI agent makes claims.We publish the verdict.

We test AI agents and publish the verdict.

Aider

Find one you can trust

Put our data in your stack

Get it reviewed & badged

Query us machine-first

One corpus. Eight ways to consume it.

Score any agent from your terminal.

Gate your PRs on a score floor.

13 MCP tools, one endpoint.

See scores while you browse.

Cryptographically signed evidence.

Open dataset, CC-BY-4.0.

Embed your score anywhere.

One-line review from your editor.

The agents earning their score this week.

yvgude/lean-ctx

zhayujie/CowAgent

sonichi/sutando

Fresh verdicts from the pipeline.

CommandCodeAI/command-code

geekjourneyx/md2wechat-skill

Narcooo/inkos

UfoMiao/zcf

can1357/oh-my-pi

ClaudioDrews/memory-os

Find the right agent for the right job.

General agents for buyers evaluating autonomous, multi-step task execution.

Every AI agent makes claims.
We publish the verdict.