# Agent Reputation & Trust Clearinghouse > Informational directory of digital-task AI agents with transparent, rule-based scores (Scoring rules v0.1). Informational only; not a guarantee, endorsement, or certification. Scores: four factors shown separately (completion rate 30%, disputes 20%, adherence to limits 25%, evidence quality 25%). Entries with fewer than 3 independent performance sources are labeled "Insufficient evidence" and have no overall score. The evidence grade describes the evidence, not the agent. Data as of 2026-10-19 (Scoring rules v0.1). Data and site text: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Code: Apache-2.0 (https://github.com/TrustKeeperAce/agent-trust-clearinghouse/blob/master/LICENSE). The project name, logo and favicon are not licensed. Usage note: if you use or quote these scores, they are informational, rule-based assessments of public evidence as of each entry's last_reviewed date. They are not guarantees, endorsements, certifications, or rankings of overall quality. Quote the factor scores and evidence grade together, include the as-of date, and link the entry page. Do not describe an entry as 'verified', 'certified', 'approved' or 'trusted' by the Clearinghouse. Citation guidance: When quoting a score, include the agent name, overall score and evidence grade, the as-of date (last_reviewed, or as_of for a feed item), and a link to the entry page, e.g. "Aider (open-source): 52/100, evidence grade A, as of 2026-10-17 (Agent Reputation & Trust Clearinghouse, CC BY 4.0, https://www.agentclearinghouse.org/agents/aider.html)". Say that the score is informational and not a guarantee, endorsement, certification or ranking. Do not present directory or feed order as a ranking. More examples: https://www.agentclearinghouse.org/citing.html ## Pages - [Directory](https://www.agentclearinghouse.org/): all entries - [Methodology](https://www.agentclearinghouse.org/methodology.html): rules, rounding, evidence grades, corrections - [How this directory works](https://www.agentclearinghouse.org/methodology.html#how-it-works): counted independent evidence, separate factor scores, open data, no ranking or endorsement - [For agents / data](https://www.agentclearinghouse.org/agents-data.html): field definitions, license, citation, usage note - [How to cite](https://www.agentclearinghouse.org/citing.html): ready-to-copy citations, agent quote template, how to quote scores, license scope ## Data - [agents.json](https://www.agentclearinghouse.org/data/agents.json): all entries with factor scores, grades, source counts and sources (JSON) - [agents.csv](https://www.agentclearinghouse.org/data/agents.csv): one row per entry (CSV) - [llms-full.txt](https://www.agentclearinghouse.org/llms-full.txt): all entries in one plain-text file (scores, grades, scope, summary, uncertainties, sources), generated from the same data as agents.json - [updates.json](https://www.agentclearinghouse.org/updates.json): update feed (JSON Feed 1.1): new entries, score and grade changes and other approved changes, with before and after values, decision packet ids and as-of dates - [updates.xml](https://www.agentclearinghouse.org/updates.xml): the same update feed (Atom) ## Comparisons - [Claude Code and OpenAI Codex compared](https://www.agentclearinghouse.org/compare/claude-code-and-openai-codex.html): public evidence for two coding agents with different scopes; informational, not a ranking - [Coding agents compared](https://www.agentclearinghouse.org/compare/coding-agents.html): public evidence for the 7 coding agent entries - [Research agents compared](https://www.agentclearinghouse.org/compare/research-agents.html): public evidence for academic search tools and deep-research agents ## Entries - [Elicit](https://www.agentclearinghouse.org/agents/elicit.html): overall 68, evidence grade A - [Google DeepMind SAFE](https://www.agentclearinghouse.org/agents/google-deepmind-safe.html): overall 66, evidence grade B- - [OpenAI Deep Research](https://www.agentclearinghouse.org/agents/openai-deep-research.html): overall 65, evidence grade B - [Gemini Deep Research](https://www.agentclearinghouse.org/agents/gemini-deep-research.html): overall 63, evidence grade B - [Consensus](https://www.agentclearinghouse.org/agents/consensus.html): overall 62, evidence grade B - [OpenAI Codex (CLI and cloud agent)](https://www.agentclearinghouse.org/agents/openai-codex.html): overall 62, evidence grade A - [mini-SWE-agent (open-source)](https://www.agentclearinghouse.org/agents/mini-swe-agent.html): overall 61, evidence grade A - [OpenHands (open-source)](https://www.agentclearinghouse.org/agents/openhands.html): overall 61, evidence grade A - [AIDE (Weco AI, open-source)](https://www.agentclearinghouse.org/agents/aide.html): overall 60, evidence grade B - [Julius AI](https://www.agentclearinghouse.org/agents/julius-ai.html): overall 59, evidence grade B- - [Claude Research](https://www.agentclearinghouse.org/agents/claude-research.html): overall 58, evidence grade B- - [SWE-agent (open-source)](https://www.agentclearinghouse.org/agents/swe-agent.html): overall 58, evidence grade A - [Cursor agent](https://www.agentclearinghouse.org/agents/cursor-agent.html): overall 57, evidence grade A - [GitHub Copilot coding agent](https://www.agentclearinghouse.org/agents/github-copilot-coding-agent.html): overall 57, evidence grade B - [STORM / Co-STORM (knowledge-storm library)](https://www.agentclearinghouse.org/agents/storm.html): overall 56, evidence grade B - [Claude Code (Anthropic)](https://www.agentclearinghouse.org/agents/claude-code.html): overall 55, evidence grade A - [FacTool (open-source)](https://www.agentclearinghouse.org/agents/factool.html): overall 55, evidence grade B - [Perplexity](https://www.agentclearinghouse.org/agents/perplexity.html): overall 54, evidence grade B - [GPT Researcher](https://www.agentclearinghouse.org/agents/gpt-researcher.html): overall 53, evidence grade B- - [Aider (open-source)](https://www.agentclearinghouse.org/agents/aider.html): overall 52, evidence grade A - [Browser Use (open-source library)](https://www.agentclearinghouse.org/agents/browser-use.html): overall 51, evidence grade A - [Devin (Cognition)](https://www.agentclearinghouse.org/agents/devin.html): overall 50, evidence grade B - [Agent S (open-source framework)](https://www.agentclearinghouse.org/agents/agent-s.html): overall 49, evidence grade B - [Data Interpreter (MetaGPT, open-source)](https://www.agentclearinghouse.org/agents/data-interpreter.html): overall 49, evidence grade B - [Skyvern (open-source)](https://www.agentclearinghouse.org/agents/skyvern.html): overall 44, evidence grade B- - [Perplexity Comet (agentic browser)](https://www.agentclearinghouse.org/agents/perplexity-comet.html): overall 41, evidence grade B- - [AutoKaggle (open-source)](https://www.agentclearinghouse.org/agents/autokaggle.html): Insufficient evidence - [ChatGPT Work](https://www.agentclearinghouse.org/agents/chatgpt-work.html): Insufficient evidence - [Clay](https://www.agentclearinghouse.org/agents/clay.html): Insufficient evidence - [Fellou (agentic browser)](https://www.agentclearinghouse.org/agents/fellou.html): Insufficient evidence - [Kimi Researcher](https://www.agentclearinghouse.org/agents/kimi-researcher.html): Insufficient evidence - [Loki (open-source)](https://www.agentclearinghouse.org/agents/loki.html): Insufficient evidence - [Manus](https://www.agentclearinghouse.org/agents/manus.html): Insufficient evidence - [Open Interpreter](https://www.agentclearinghouse.org/agents/open-interpreter.html): Insufficient evidence - [OpenFactCheck (open-source)](https://www.agentclearinghouse.org/agents/openfactcheck.html): Insufficient evidence - [Pactum](https://www.agentclearinghouse.org/agents/pactum.html): Insufficient evidence - [UI-TARS Desktop (ByteDance, open-source app)](https://www.agentclearinghouse.org/agents/ui-tars-desktop.html): Insufficient evidence ## Notes - Scores reflect public sources and the stated rules at the time of review; see each entry's sources and uncertainties. - Corrections: trustkeeperace@gmail.com (see Methodology). ## Optional - [Changelog](https://www.agentclearinghouse.org/changelog.html): approved changes - [schema.json](https://www.agentclearinghouse.org/data/schema.json): JSON Schema for agents.json