Skip to content

docs: map user goals to CLI commands and outcomes - #48

Merged
laithalsaadoon merged 2 commits into
mainfrom
factory/13-readme-matrix
Oct 9, 2026
Merged

laithalsaadoon merged 2 commits into
mainfrom
factory/13-readme-matrix

Conversation

@laithalsaadoon

@laithalsaadoon laithalsaadoon commented Oct 9, 2026 •

Copy link
Copy Markdown
Owner

The README maps reader goals to commands and useful outcomes, so readers can find prior work by topic, compare Claude Code and Codex histories, inspect recurring corrections, and keep their history current.

The matrix covers all registered CLI commands and links each to its reference. It distinguishes previews from paid Bedrock execution, explains search prerequisites and lower-bound cost estimates, and makes scheduled embedding and analysis costs explicit. The CLI reference adds the lake commands and cross-corpus flags. A drift test checks command coverage and reference anchors.

Validation, checked 2026-10-09:

  • mise run check: passed, 2,490 tests passed, 1 skipped, 1 expected failure.
  • mise run docs:gate: passed, 73 site tests and 1,962 internal links with 0 broken.
  • Command-matrix tests: 8 passed; every registered command and reference anchor is covered.
  • CLI help and implementation review for command defaults and prerequisites.

The prose warning baseline falls by one warning to match the measured count.

laithalsaadoon and others added 2 commits October 9, 2026 16:36
The README listed the CLI in prose and one package-table row, so a reader
could not see search, embed, analyze, the lake and corpus groups or the
analytics views. A "What you can do" table now covers all 16 registered
commands plus three headline queries taken from `atif-sql examples`, each
linking to its docs/reference/cli.md anchor, with the Bedrock commands marked
as costing money.

Corrections the CLI backs: the atif-analytics row named trajectory, cluster,
terms and community pipelines that `analyze` does not run (it runs classify,
conflicts, friction and perceived), and the Install note called search
dry-run by default when only analyze is (embed needs --limit or --all,
search embeds one query per call). cli.md gains the missing `lake` section
and the --all-corpora and --no-lake flags on query and search. Quick start
prose the matrix now carries is trimmed.

test_readme_matrix.py walks the cyclopts command tree and fails when a
command has no row, when a row names a command the app lacks, when a link
misses its cli.md heading, and when the matrix parses zero rows.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Readers choose commands to recover prior work, compare agent histories,
review recurring corrections, and keep their session history current.
Describe those goals and their outcomes before the implementation details.

Clarify paid execution defaults, conversion output, search prerequisites,
lower-bound cost estimates, and paid scheduled embedding and analysis.
Lower the existing prose baseline to its observed warning count.
@laithalsaadoon laithalsaadoon changed the title docs: add a task, command, outcome matrix to the README docs: map user goals to CLI commands and outcomes Oct 9, 2026
@laithalsaadoon
laithalsaadoon merged commit 7ea99bb into main Oct 9, 2026
18 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant