Directories
¶
| Path | Synopsis |
|---|---|
|
eval
|
|
|
agent
Package agent defines the interface every coding-agent CLI is driven through.
|
Package agent defines the interface every coding-agent CLI is driven through. |
|
derive
Package derive builds an eval suite for a skill that has none, by recovering a spec from the bundle and drafting evals from it.
|
Package derive builds an eval suite for a skill that has none, by recovering a spec from the bundle and drafting evals from it. |
|
lint
Package lint validates a skill bundle across three layers: spec conformance, quality, and injection (security scan).
|
Package lint validates a skill bundle across three layers: spec conformance, quality, and injection (security scan). |
|
llm
Package llm is the single seam through which every model call passes.
|
Package llm is the single seam through which every model call passes. |
|
llm/agentcli
Package agentcli implements the LLM gateway by shelling out to a coding-agent CLI in print mode, billing to the operator's existing subscription.
|
Package agentcli implements the LLM gateway by shelling out to a coding-agent CLI in print mode, billing to the operator's existing subscription. |
|
llm/api
Package api implements the LLM gateway against a direct Anthropic-compatible HTTP endpoint.
|
Package api implements the LLM gateway against a direct Anthropic-compatible HTTP endpoint. |
|
llm/fake
Package fake provides an in-memory Gateway for tests.
|
Package fake provides an in-memory Gateway for tests. |
|
provider
Package provider is the one place the environment is mapped onto an LLM backend.
|
Package provider is the one place the environment is mapped onto an LLM backend. |
|
report
Package report defines the machine-readable result of an eval run: the schema a CI job reads, the worker posts, and the UI charts.
|
Package report defines the machine-readable result of an eval run: the schema a CI job reads, the worker posts, and the UI charts. |
|
runner
Package runner plans and executes an evaluation: which sessions to run, in what environment, with what resumed from a previous attempt.
|
Package runner plans and executes an evaluation: which sessions to run, in what environment, with what resumed from a previous attempt. |
|
sandbox
Package sandbox is the seam every evaluation run executes through.
|
Package sandbox is the seam every evaluation run executes through. |
|
sandbox/docker
Package docker implements the sandbox driver on the Docker CLI.
|
Package docker implements the sandbox driver on the Docker CLI. |
|
sandbox/imagespec
Package imagespec renders everything an image needs as text: the base Dockerfile, a per-skill layer over it, the proxy configuration that enforces a network allowlist, and the digest that makes a layer cacheable.
|
Package imagespec renders everything an image needs as text: the base Dockerfile, a per-skill layer over it, the proxy configuration that enforces a network allowlist, and the digest that makes a layer cacheable. |
|
spec
Package spec holds the SkillSpec IR that generation, suite drafting, and scoring all derive from.
|
Package spec holds the SkillSpec IR that generation, suite drafting, and scoring all derive from. |
|
store
Package store owns the .whetstone/ workspace: SQLite index, completion cache, and the on-disk artifact layout.
|
Package store owns the .whetstone/ workspace: SQLite index, completion cache, and the on-disk artifact layout. |
|
suite
Package suite is a skill's eval set: the evals/evals.json tasks it is measured against, the evals/triggers.json queries its description is measured against, and a shipped distractor pack that makes a trigger measurement discriminating.
|
Package suite is a skill's eval set: the evals/evals.json tasks it is measured against, the evals/triggers.json queries its description is measured against, and a shipped distractor pack that makes a trigger measurement discriminating. |
|
suite/assets
Package assets embeds the static suite assets.
|
Package assets embeds the static suite assets. |
|
tune
Package tune optimizes a skill's description for trigger accuracy.
|
Package tune optimizes a skill's description for trigger accuracy. |
|
Package evalqueue is the eval job queue, backed by Postgres claim/lease/heartbeat.
|
Package evalqueue is the eval job queue, backed by Postgres claim/lease/heartbeat. |
|
Package evalsuite is the registry-side counterpart to internal/eval/suite: evaluation task-suites become content-addressable artifacts stored alongside skill bundles, so a quality score can be re-run later against the exact same tasks with a different model panel.
|
Package evalsuite is the registry-side counterpart to internal/eval/suite: evaluation task-suites become content-addressable artifacts stored alongside skill bundles, so a quality score can be re-run later against the exact same tasks with a different model panel. |
|
Package gate decides whether a skill version may be published, held for review, or refused outright.
|
Package gate decides whether a skill version may be published, held for review, or refused outright. |
|
Package ownership resolves who owns a skill name and who may manage a pattern.
|
Package ownership resolves who owns a skill name and who may manage a pattern. |
|
Package quality stores and retrieves per-version quality scores.
|
Package quality stores and retrieves per-version quality scores. |
|
Package worker implements the eval queue worker's run loop: claim, materialise, evaluate, report.
|
Package worker implements the eval queue worker's run loop: claim, materialise, evaluate, report. |
Click to show internal directories.
Click to hide internal directories.