feat(graphile-llm): wire billing metering, inference logging, and global routes#1192

Merged

pyramation merged 21 commits into

mainfrom

feat/llm-billing-metering

May 21, 2026

Commits on May 20, 2026

feat(graphile-llm): wire billing metering into LLM plugins

pyramation
authored and
devin-ai-integration[bot]
committed
refactor(graphile-llm): separate metering into standalone LlmMeteringPlugin

pyramation
authored and
devin-ai-integration[bot]
committed
feat(graphile-llm): meter slug = model name, three-level waterfall

pyramation
authored and
devin-ai-integration[bot]
committed
remove docs/spec/llm-metering.md (moving to constructive-db)

pyramation
authored and
devin-ai-integration[bot]
committed
fix(graphile-llm): rename encrypted_secrets_module to config_secrets_user_module

pyramation
authored and
devin-ai-integration[bot]
committed
fix(graphile-llm): remove estimatedEmbeddingTokens, estimate from text length

pyramation
authored and
devin-ai-integration[bot]
committed
feat(graphile-llm): add agent discovery plugin and REST streaming endpoint

pyramation
authored and
devin-ai-integration[bot]
committed
fix: restore inline documentation comments removed in previous commits

pyramation
authored and
devin-ai-integration[bot]
committed
fix: restore remaining inline comments (inject/remove, recurse, replace, where/filter)

pyramation
authored and
devin-ai-integration[bot]
committed
refactor: remove fake token estimation from metering

pyramation
authored and
devin-ai-integration[bot]
committed
refactor(graphile-llm): centralize LLM env config in env.ts

pyramation
authored and
devin-ai-integration[bot]
committed
refactor: replace smart tag discovery with module config query

pyramation
authored and
devin-ai-integration[bot]
committed
refactor: remove apiKey/api_key_ref from graphile-llm
pyramation
committed
feat(llm-api): wire metering + add global /v1/threads routes
pyramation
committed
feat(llm-api): add inference logging to usage_log_inference table
pyramation
committed
test(llm-api): add unit tests for metering + route registration
pyramation
committed
refactor(llm-api): use generate() only, remove generateWithUsage runtime check
pyramation
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat(graphile-llm): wire billing metering, inference logging, and global routes#1192

feat(graphile-llm): wire billing metering, inference logging, and global routes#1192
pyramation merged 21 commits into
mainfrom
feat/llm-billing-metering

Commits on May 20, 2026

feat(graphile-llm): wire billing metering into LLM plugins

refactor(graphile-llm): separate metering into standalone LlmMeteringPlugin

feat(graphile-llm): meter slug = model name, three-level waterfall

remove docs/spec/llm-metering.md (moving to constructive-db)

fix(graphile-llm): rename encrypted_secrets_module to config_secrets_user_module

fix(graphile-llm): remove estimatedEmbeddingTokens, estimate from text length

feat(graphile-llm): add agent discovery plugin and REST streaming endpoint

fix: restore inline documentation comments removed in previous commits

fix: restore remaining inline comments (inject/remove, recurse, replace, where/filter)

refactor: remove fake token estimation from metering

refactor(graphile-llm): centralize LLM env config in env.ts

refactor: replace smart tag discovery with module config query

refactor: remove apiKey/api_key_ref from graphile-llm

feat(llm-api): wire metering + add global /v1/threads routes

feat(llm-api): add inference logging to usage_log_inference table

test(llm-api): add unit tests for metering + route registration

refactor(llm-api): use generate() only, remove generateWithUsage runtime check

Commits on May 21, 2026

fix(graphile-llm): restore embed latency timing in text-mutation-plugin

fix(graphile-llm): add latency timing to search embed and unmetered embed paths

refactor(graphile-cache): add ModuleConfigCache<T>, migrate all LLM plugin caches

refactor(llm-api): replace token_usage jsonb with model text on message INSERT