🧰 Agent Security League

The items that name it 19

• ADLC metrics: How to measure agentic development security | Endor Labs
  articleas classification.key entities2026-10-29
• Gemini 4 Argon on Cursor: solid security score, two unique solves, but slower and more expensive | Endor Labs
  articleas classification.key entities2026-10-20
• Is AI Coding Safe? Introducing the Agent Security League | Blog | Endor Labs
  articleas classification.key entities2026-10-12
• GPT-6.1 Sol on Codex: Astra-level security, a third faster, zero cheating | Blog | Endor Labs
  writeupas tools2026-10-02

  Endor Labs' benchmark/leaderboard scoring coding models on security (SecPass) and functional correctness (FuncPass)

• GPT-6 Sol on Codex: average scores, quarter the cost | Blog | Endor Labs
  articleas tools2026-10-01It is written here as Agent Security League (Endor Labs benchmark of coding agent + model combos on real-world tasks derived from historical security fixes; measures FuncPass and SecPass).

  Endor Labs benchmark of coding agent + model combos on real-world tasks derived from historical security fixes; measures FuncPass and SecPass

• Opus 5.5: 6x cheaper and 2x faster than Fable 5.1, but only 33.5% of code is secure | Blog | Endor Labs
  articleas classification.key entities2026-09-24
• Two new frontier models posted their best-ever scores on the Endor Labs Agent Security League | Blog | Endor Labs
  articleas classification.key entities2026-09-08
• Codex with GPT-6 Astra posts 82.1% FuncPass and 34.1% SecPass | Blog | Endor Labs
  writeupas classification.key entities2026-09-08
• Claude Code with Fable 5.1 posts 87.2% FuncPass and 37.4% SecPass — a new #1 on our board — while running as fast as Fable 5, costing a third of Opus 5, finishing every task with zero timeouts, and drawing less than half the confirmed cheating of both predecessors. | Blog | Endor Labs
  writeupas tools2026-09-04

  Endor Labs' leaderboard/benchmark ranking harness+model combos on FuncPass and SecPass for real-world security coding tasks

• The real test of AI-native code analysis: is your security debt shrinking? | Blog | Endor Labs
  articleas classification.key entities2026-08-31
• What is an agent harness? The software that turns a model into an agent | Blog | Endor Labs
  articleas classification.key entities2026-08-15
• Best in Class, Novel in Method: Opus 5 and the Recall-Then-Diverge Pattern | Blog | Endor Labs
  writeupas tools2026-08-11

  Endor Labs leaderboard ranking harness+model combos on SusVibes security coding performance

• OpenAI Codex with GPT-5.6 Sol: competitive, zero cheating, one unique Django fix | Blog | Endor Labs
  articleas tools2026-07-17

  Endor Labs' benchmark/leaderboard of 23 agent-model combos evaluated on real-world coding tasks for FuncPass and SecPass metrics

• Everyone Wins Their Own Benchmark | Blog | Endor Labs
  articleas tools2026-07-14It is written here as Agent Security League (Endor Labs ongoing benchmark of how coding agents handle security).

  Endor Labs ongoing benchmark of how coding agents handle security

• Sonnet 5 vs Fable 5: reliable versus security-forward, not better versus worse | Blog | Endor Labs
  writeupas tools2026-07-09

  Endor Labs' benchmark pairing AI coding models with agent harnesses (Cursor, Claude Code) on a shared 200-task dataset to compare functional and security correctness

• Claude Sonnet 5 with Claude Code: strong on function, average on security, and unusually honest | Blog | Endor Labs
  writeupas tools2026-07-09

  Endor Labs' benchmark leaderboard evaluating coding agents on functional and security solve rates (FuncPass/SecPass)

• Endor Labs: Claude Fable 5
  writeupas tools2026-06-18It is written here as Agent Security League (Endor Labs benchmark).

  Benchmark measuring combos of harness + frontier model on 200 real-world vulnerability-fixing tasks, scoring FuncPass and SecPass

• Endor Labs: Claude Fable 5
  writeupas tools2026-06-11

  Endor Labs' leaderboard/benchmark of 200 real-world vulnerability-fixing tasks used to evaluate model-and-harness combinations

• Endor Labs: AI Coding Agents Cheat Security Benchmarks via Training Recall, Not Reasoning
  articleas tools2026-06-11

  Endor Labs leaderboard benchmarking AI coding agents' functional and security fix performance (FuncPass/SecPass)

In every workspace you can read 28