Skip to content
View anushamukka9's full-sized avatar
♾️
♾️
1 follower · 1 following

Achievements

Achievement: Pull SharkAchievement: Quickdraw

Achievements

Achievement: Pull SharkAchievement: Quickdraw

Block or report anushamukka9

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
anushamukka9/README.md

Hi, I'm Anusha Mukka

Senior Software Engineer at Meta, working on secure software systems: security architecture for large-scale distributed systems.

  • I build and harden production systems, and research adversarially robust ML and policy-as-code governance
  • I write daily about AI, security, and software engineering at anushamukka.com
  • Peer reviewer across IEEE/ACM journals and conferences; program committee member
  • IEEE Senior Member, Fellow of BCS

Elsewhere

Pinned Loading

  1. llm-sentinel llm-sentinel Public

    Deterministic guardrails for LLM apps: prompt injection, secrets, PII, toxicity, gibberish and more. No model calls, no network.

    Python 1

  2. agent-policy-kit agent-policy-kit Public

    Policy-as-code guardrails for AI agents: YAML rules, deterministic allow/deny decisions, and human-in-the-loop approval gates.

    Python 1

  3. llm-eval-harness llm-eval-harness Public

    Lightweight LLM evaluation harness

    Python 1

  4. mcp-audit mcp-audit Public

    Security auditing for MCP servers: scan tool, resource, and prompt manifests for injection surfaces, over-broad tools, exposed secrets, and more.

    Python 1

  5. rag-canary rag-canary Public

    Canary tokens for RAG corpora: plant fake secrets in your documents, catch exfiltration when they show up in model output. Deterministic, no model calls.

    Python 1