code-lens-cli 0.9.1__tar.gz → 0.9.2__tar.gz

{code_lens_cli-0.9.1 → code_lens_cli-0.9.2}/CHANGELOG.md

@@ -5,6 +5,13 @@ All notable changes to this project will be documented in this file.
 Format follows [Keep a Changelog](https://keepachangelog.com/). This project
 adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [0.9.2] - 2026-05-17
+
+### Changed
+
+- CLAUDE.md + README.md: add Name section explaining antoine = N to 1 (collapse N ad-hoc tool calls into one `kata` / `antoine` verb; `kata-cli` is the PyPI distribution label, not a command name).
+- README.md: expand "What's here" to cover the three things antoine manages — the CLI surface (dual-published as `antoine-cli` / `kata-cli` / `code-lens-cli`), the `experiments/scripts_eval/` A/B harness, and the recorded results in `docs/eval-rounds/`.
+
 ## [0.9.1] - 2026-05-17
 
 ### Changed

{code_lens_cli-0.9.1 → code_lens_cli-0.9.2}/CLAUDE.md

@@ -2,6 +2,20 @@
 
 This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
 
+## Name
+
+**antoine** = **"N to 1"** (an-to-one). The name encodes the goal: collapse
+the N ad-hoc tool calls an agent would otherwise make against a codebase
+(`ls` + `cat` + `grep` + `git log` + `git show` + …) into **one** call to a
+purpose-built `kata` verb (or its `antoine` alias — both console scripts are
+defined in [`pyproject.toml`](./pyproject.toml); `kata-cli` is the PyPI
+**distribution** name, not a command name) that returns the same information
+as structured data. Every verb antoine ships is a bet that some recurring
+N-call pattern
+has a 1-call replacement that is cheaper, more reliable, and easier to
+delegate to a subagent. You are that agent — when you face a task shape
+covered by the dispatching table below, prefer the 1-call verb.
+
 ## Project Status
 
 `antoine` is an AgentCulture sibling repo — **codebase lookup and indexing for

code_lens_cli-0.9.2/PKG-INFO

@@ -0,0 +1,71 @@
+Metadata-Version: 2.4
+Name: code-lens-cli
+Version: 0.9.2
+Summary: antoine — codebase lookup and indexing for agent skills (greenfield AgentCulture sibling).
+Project-URL: Homepage, https://github.com/agentculture/antoine
+Project-URL: Issues, https://github.com/agentculture/antoine/issues
+Author: AgentCulture
+License-Expression: MIT
+License-File: LICENSE
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Software Development
+Requires-Python: >=3.12
+Requires-Dist: pyyaml>=6.0
+Description-Content-Type: text/markdown
+
+# antoine
+
+Codebase lookup and indexing for agent skills.
+
+The name encodes the goal: **antoine = "N to 1"** (an-to-one). Collapse the
+N ad-hoc tool calls an agent would otherwise make against a codebase
+(`ls` + `cat` + `grep` + `git log` + `git show` + …) into **one** call to a
+purpose-built `kata` verb (or its `antoine` alias — see
+[`pyproject.toml`](./pyproject.toml) for the console-script wiring;
+`kata-cli` is the PyPI **distribution** name, not a command name) that
+returns the same information as structured data. Every verb antoine ships is a bet that some recurring
+N-call pattern has a 1-call replacement that is cheaper, more reliable,
+and easier to delegate to a subagent.
+
+## What's here
+
+This repo manages three intertwined things — the CLI that ships the
+1-call verbs, the evaluation harness that tells us whether the verbs are
+actually worth the bet, and the recorded results from past rounds.
+
+- **`antoine/`** — the package that will eventually expose the lookup
+  verbs. Greenfield: `learn` / `explain` / `whoami` are honest
+  placeholder stubs. See [`CLAUDE.md`](./CLAUDE.md) for build / test /
+  architecture details.
+
+- **`kata-cli` (and `code-lens-cli`) — alt-published PyPI distributions**
+  carrying the same wheel content as `antoine-cli`. Installing any of the
+  three exposes the same pair of console scripts — `antoine` and `kata`
+  (see [`pyproject.toml`](./pyproject.toml)). The dispatching directives
+  baked into [`CLAUDE.md`](./CLAUDE.md) refer to **verbs** invoked via the
+  `kata` / `antoine` commands; `kata-cli` is the distribution label users
+  `pip install`, not a command they run. The dual-publish loop is defined
+  in [`.github/workflows/`](./.github/workflows/); see
+  [`CHANGELOG.md`](./CHANGELOG.md) entries for v0.7.0 / v0.7.1 for the
+  history of how the three distribution names were wired up.
+
+- **`experiments/scripts_eval/`** — the A/B-test harness for the
+  `repo-map` skill (env-var-gated hooks, three-layer scoring, 5-repo
+  round-1 corpus, multi-arm rider design with banned / directed /
+  organic modes, multi-pair LLM-as-judge). The eval rounds are the
+  validation gate before the `learn` / `explain` / `overview` /
+  `doctor` verb redesign lands. See
+  [`experiments/scripts_eval/README.md`](./experiments/scripts_eval/README.md)
+  and
+  [`experiments/scripts_eval/RUNBOOK.md`](./experiments/scripts_eval/RUNBOOK.md).
+
+- **`docs/eval-rounds/`** — write-ups from completed evaluation rounds
+  (round 01, smoke 02, round 02 so far). These are the empirical record
+  behind the directives in [`CLAUDE.md`](./CLAUDE.md) — including the
+  round-2 finding that subagents build their plans from the prompt body
+  *before* consulting the skills catalog, which is why the dispatching
+  table lives in the parent agent's instructions rather than in skill
+  descriptions.

code_lens_cli-0.9.2/README.md

@@ -0,0 +1,53 @@
+# antoine
+
+Codebase lookup and indexing for agent skills.
+
+The name encodes the goal: **antoine = "N to 1"** (an-to-one). Collapse the
+N ad-hoc tool calls an agent would otherwise make against a codebase
+(`ls` + `cat` + `grep` + `git log` + `git show` + …) into **one** call to a
+purpose-built `kata` verb (or its `antoine` alias — see
+[`pyproject.toml`](./pyproject.toml) for the console-script wiring;
+`kata-cli` is the PyPI **distribution** name, not a command name) that
+returns the same information as structured data. Every verb antoine ships is a bet that some recurring
+N-call pattern has a 1-call replacement that is cheaper, more reliable,
+and easier to delegate to a subagent.
+
+## What's here
+
+This repo manages three intertwined things — the CLI that ships the
+1-call verbs, the evaluation harness that tells us whether the verbs are
+actually worth the bet, and the recorded results from past rounds.
+
+- **`antoine/`** — the package that will eventually expose the lookup
+  verbs. Greenfield: `learn` / `explain` / `whoami` are honest
+  placeholder stubs. See [`CLAUDE.md`](./CLAUDE.md) for build / test /
+  architecture details.
+
+- **`kata-cli` (and `code-lens-cli`) — alt-published PyPI distributions**
+  carrying the same wheel content as `antoine-cli`. Installing any of the
+  three exposes the same pair of console scripts — `antoine` and `kata`
+  (see [`pyproject.toml`](./pyproject.toml)). The dispatching directives
+  baked into [`CLAUDE.md`](./CLAUDE.md) refer to **verbs** invoked via the
+  `kata` / `antoine` commands; `kata-cli` is the distribution label users
+  `pip install`, not a command they run. The dual-publish loop is defined
+  in [`.github/workflows/`](./.github/workflows/); see
+  [`CHANGELOG.md`](./CHANGELOG.md) entries for v0.7.0 / v0.7.1 for the
+  history of how the three distribution names were wired up.
+
+- **`experiments/scripts_eval/`** — the A/B-test harness for the
+  `repo-map` skill (env-var-gated hooks, three-layer scoring, 5-repo
+  round-1 corpus, multi-arm rider design with banned / directed /
+  organic modes, multi-pair LLM-as-judge). The eval rounds are the
+  validation gate before the `learn` / `explain` / `overview` /
+  `doctor` verb redesign lands. See
+  [`experiments/scripts_eval/README.md`](./experiments/scripts_eval/README.md)
+  and
+  [`experiments/scripts_eval/RUNBOOK.md`](./experiments/scripts_eval/RUNBOOK.md).
+
+- **`docs/eval-rounds/`** — write-ups from completed evaluation rounds
+  (round 01, smoke 02, round 02 so far). These are the empirical record
+  behind the directives in [`CLAUDE.md`](./CLAUDE.md) — including the
+  round-2 finding that subagents build their plans from the prompt body
+  *before* consulting the skills catalog, which is why the dispatching
+  table lives in the parent agent's instructions rather than in skill
+  descriptions.

{code_lens_cli-0.9.1 → code_lens_cli-0.9.2}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "code-lens-cli"
-version = "0.9.1"
+version = "0.9.2"
 description = "antoine — codebase lookup and indexing for agent skills (greenfield AgentCulture sibling)."
 readme = "README.md"
 license = "MIT"

{code_lens_cli-0.9.1 → code_lens_cli-0.9.2}/uv.lock

@@ -32,7 +32,7 @@ wheels = [
 
 [[package]]
 name = "antoine-cli"
-version = "0.9.1"
+version = "0.9.2"
 source = { editable = "." }
 dependencies = [
     { name = "pyyaml" },

code_lens_cli-0.9.1/PKG-INFO

@@ -1,36 +0,0 @@
-Metadata-Version: 2.4
-Name: code-lens-cli
-Version: 0.9.1
-Summary: antoine — codebase lookup and indexing for agent skills (greenfield AgentCulture sibling).
-Project-URL: Homepage, https://github.com/agentculture/antoine
-Project-URL: Issues, https://github.com/agentculture/antoine/issues
-Author: AgentCulture
-License-Expression: MIT
-License-File: LICENSE
-Classifier: Development Status :: 3 - Alpha
-Classifier: Intended Audience :: Developers
-Classifier: License :: OSI Approved :: MIT License
-Classifier: Programming Language :: Python :: 3.12
-Classifier: Topic :: Software Development
-Requires-Python: >=3.12
-Requires-Dist: pyyaml>=6.0
-Description-Content-Type: text/markdown
-
-# antoine
-
-Codebase lookup and indexing for agent skills.
-
-## What's here
-
-- **`antoine/`** — the package that will eventually expose the lookup
-  verbs. Greenfield: `learn` / `explain` / `whoami` are honest
-  placeholder stubs. See [`CLAUDE.md`](./CLAUDE.md) for build / test /
-  architecture details.
-
-- **`experiments/scripts_eval/`** — A/B-test harness for the `repo-map`
-  skill (env-var-gated hooks, three-layer scoring, 5-repo round-1
-  corpus). Round 1 is the validation gate before the verb design
-  lands. See
-  [`experiments/scripts_eval/README.md`](./experiments/scripts_eval/README.md)
-  and
-  [`experiments/scripts_eval/RUNBOOK.md`](./experiments/scripts_eval/RUNBOOK.md).

code_lens_cli-0.9.1/README.md

@@ -1,18 +0,0 @@
-# antoine
-
-Codebase lookup and indexing for agent skills.
-
-## What's here
-
-- **`antoine/`** — the package that will eventually expose the lookup
-  verbs. Greenfield: `learn` / `explain` / `whoami` are honest
-  placeholder stubs. See [`CLAUDE.md`](./CLAUDE.md) for build / test /
-  architecture details.
-
-- **`experiments/scripts_eval/`** — A/B-test harness for the `repo-map`
-  skill (env-var-gated hooks, three-layer scoring, 5-repo round-1
-  corpus). Round 1 is the validation gate before the verb design
-  lands. See
-  [`experiments/scripts_eval/README.md`](./experiments/scripts_eval/README.md)
-  and
-  [`experiments/scripts_eval/RUNBOOK.md`](./experiments/scripts_eval/RUNBOOK.md).