Superagent
6,884Security layer protecting AI apps from prompt injection, data leaks, and harmful outputs.
Activity band details
Activity band A to F: a bin of a score that adds GitHub stars (log scale), a push to the upstream repository in the 12 months before `as_of`, the number of declared integration surfaces and, where the catalogue records them, citing papers and benchmark evaluations. It measures popularity and activity. It is not a level of evidence: no band means a paper cites the tool or that it was measured. Recency is judged as of 2026-11-22.
Computed band: D
(raw score 4.54).
Shown band: D
.
Facets:
documented_in=0evaluated_on=0integration_surfaces=1paper_citations=0recent_push=2stars=1.54
Source of truth: pipeline/maturity/score.py
Description
A security wrapper that protects AI applications against prompt injections, data leaks, and harmful outputs. Developers embed this safety layer directly into their applications to enforce runtime security policies and demonstrate compliance.
Why it's included
Occupies prevention × in_generation × fully_autonomous as an embedded security wrapper: developers link the library directly into the application so policy enforcement runs inline with the LLM call rather than behind a network proxy or as a post-hoc validator (contrast guardrails-ai). Emphasizes compliance-ready runtime enforcement, making it a useful exemplar of in-process guardrail deployment.
Addresses failure modes
These tags are the catalog's claim about what Superagent is built to catch or prevent. A tag is not a test result. SAICA-Bench has not measured Superagent, so none of these tags is tested here.
- Supply-chain attack — Slopsquatting, typosquatting, compromised MCP servers.