Research

Following opens with early access — sign up for early access to register your interest; it does not switch following on today. Everything you follow will land on one page, in your account menu.

1. Dev Tooling & Infra✓ verified

   Researcher shows one malicious GitHub issue could hijack repos running Claude Code's GitHub Action

   CI/CD-embedded coding agents inherit the write access of the workflow they run in — treat any agent-triggering input (issue titles, PR bodies, comments) from an untrusted user as untrusted, patched or not.

   GMO Flatt Security (independent research)·2026-06-01
2. Dev Tooling & Infra✓ verified

   GitHub Copilot moves every plan to token-metered 'AI Credits'

   Usage-metered agent tooling makes cost track how hard your agents actually work — the same budgeting shift teams hit running their own multi-agent lanes.

   GitHub·2026-06-01
3. Dev Tooling & Infra✓ verified

   GitHub lets you assign a dependency alert straight to an AI agent to fix

   Dependency triage is fatigue-heavy toil an agent can genuinely take off your plate — as long as the merge stays a human decision.

   GitHub·2026-04-07
4. Dev Tooling & Infra✓ verified

   A poisoned npm package quietly rewrote a coding agent's memory — and it reloaded every session

   Treat any automatic edit to an agent's memory or instruction files as a reviewable event, not a silent auto-load — a single poisoned dependency can otherwise steer every future run.

   Cisco·2026-04-01
5. Dev Tooling & Infra✓ verified

   mem0 / agent-memory architectures maturing

   AI Uni's own memory-architecture work sits in this fast-moving category — the anti-evaporation problem the platform substrate solves.

   mem0·2026-04-01
6. Dev Tooling & Infra

   KV-cache agent-state persistence: reported 89% better completion, 67% fewer calls

   If the effect holds, strong evidence for on-disk substrate-loading + persistence work. UNVERIFIED effect size — find the primary benchmark before citing the numbers.

   mem0 (vendor blog)·2026-04-01
7. Dev Tooling & Infra✓ verified

   Agent observability field: LangSmith vs Braintrust vs Langfuse vs Arize

   Informs deterministic-vs-LLM-judge layering. Braintrust's merge-blocking eval-action is a pattern to study — but LLM-judge stays post-hoc, never replacing deterministic CI gates.

   Braintrust / Latitude·2026-03-15
8. Dev Tooling & Infra✓ verified

   'ClawHavoc': 900+ malicious agent 'skills' pushed through a third-party marketplace

   Any skill, tool, or config you didn't write deserves the same scrutiny as a code dependency — and 'delete the skill' is not the same as 'undo what it did.'

   Koi Security·2026-02-01