Release/Versioning

Issue 91 Edition 2026-04-01 3 min read

Not accepted General

Sources: 1 • Confidence: Medium • Updated: 2026-04-12 09:58

Key takeaways

datasette-enrichments-llm version 0.2a1 has been released.
When an enrichment is triggered, the triggering actor is now passed through to the LLM call as the actor argument to llm.mode(...).

When an enrichment is triggered, the triggering actor is now passed through to the LLM call as the actor argument to llm.mode(...).

What other changes (API, behavior, configuration, breaking changes) are included in datasette-enrichments-llm 0.2a1 beyond actor propagation?
How is “actor” defined and serialized at the enrichment trigger boundary (fields, identifiers), and are there any compatibility expectations with existing Datasette actor implementations?
Are there documented security/privacy considerations for passing actor into LLM calls (e.g., logging, prompt content, data minimization)?
Is 0.2a1 intended for production use, and what is the expected stability/support window for this alpha series?
Is there any direct decision-readthrough (operator, product, or investor) in this corpus beyond the narrow integration implication of actor propagation?

Project is iterating quickly with new alpha releases, suggesting ongoing development cadence that could increase adoption among Datasette users who need LLM enrichments.
Passing actor into LLM calls indicates movement toward user aware governance such as auditing, authorization, and rate controls at invocation time.
Actor propagation may enable differentiated behavior per user or service account, implying potential for broader enterprise readiness if complemented by policy and logging features.

Release notes for 0.2a1 show additional governance or admin features such as audit logs, permissions integration, rate limiting, or configuration for actor handling.
Documentation clarifies actor schema and compatibility with common Datasette actor implementations and includes guidance for privacy and data minimization in prompts and logs.
Subsequent versions move from alpha to stable and communicate support expectations, indicating reliability focus beyond a narrow integration tweak.

0.2a1 adds little beyond actor pass through and future releases stall, implying limited scope and weak momentum for broader governance capabilities.
Actor serialization is inconsistent or incompatible across Datasette deployments, causing breakage or requiring heavy migration, reducing practical value of actor aware LLM calls.
Security or privacy concerns emerge around passing actor data into LLM prompts or logs without mitigations, discouraging adoption in regulated settings.