claude-code CLI

Eight principles, scored

Behavioral and source checks. Every non-pass row carries a copy-paste remediation prompt.

1. P1
   Non-Interactive by Default

   All 6 checks pass.

   pass
2. P2
   Structured, Parseable Output

   --output/--format flag detected but could not validate JSON via safe probes (--help/--version override output flags in most CLIs) · CLI emits structured output but exposes no `schema` subcommand or `--schema` flag at top level or nested one level deep. Agents need a runtime-discoverable schema to pin against shape changes.

   Remediation · P2⧉ copy prompt

   Make claude-code satisfy P2 (Structured, Parseable Output) from the agent-native CLI standard. Address:
   - Structured output support (--output/--format flag detected but could not validate JSON via safe probes (--help/--version override output flags in most CLIs))
   - Structured-output CLI exposes its schema at runtime (CLI emits structured output but exposes no `schema` subcommand or `--schema` flag at top level or nested one level deep. Agents need a runtime-discoverable schema to pin against shape changes.)
   - --json / --jsonl short aliases for --output (no --json or --jsonl short alias found. Agents and pipelines benefit from short forms alongside the canonical `--output` enum.)
   - `--raw` flag for pipe-safe unformatted output (no `--raw` flag advertised. MAY-tier — useful for pipelines that want to strip formatting before piping to other tools.)
   - Bad invocation exits with structured usage-error code (2) (bad invocation exited with code 1. The 0/1/2/77/78 convention reserves code 2 for usage errors; using a different non-zero code (often 1) blurs the distinction between usage errors and general failure.)
   - Errors emit JSON envelope with discriminant + type id + message roles under `--output json` (bad invocation under `--output json` produced no parseable JSON on stderr or stdout. JSON mode must emit a JSON error envelope, not plain text.)
   Requirements: https://anc.dev/p2

   fail
3. P3
   Progressive Help Discovery

   no `examples` subcommand or `--examples` flag found. MAY-tier — a curated usage block keeps agents from hunting through long help text. · `-h` and `--help` produce byte-identical output. SHOULD-tier — clap renders the short summary on `-h` and the full description on `--help` when `long_about` is set; collapsing them gives agents no concise list-level grep target.

   Remediation · P3⧉ copy prompt

   Make claude-code satisfy P3 (Progressive Help Discovery) from the agent-native CLI standard. Address:
   - `examples` subcommand or `--examples` flag for curated usage patterns (no `examples` subcommand or `--examples` flag found. MAY-tier — a curated usage block keeps agents from hunting through long help text.)
   - Short `-h` summary differs from `--help` long form (`-h` and `--help` produce byte-identical output. SHOULD-tier — clap renders the short summary on `-h` and the full description on `--help` when `long_about` is set; collapsing them gives agents no concise list-level grep target.)
   - Each subcommand's `--help` ships at least one invocation example (subcommands missing example invocations in their `--help`: agents, attach, auth, auto-mode, doctor, gateway, import, install, logs, purge, respawn, rm, setup-token, ultrareview. Examples teach agents the call shape faster than option tables; use clap's `after_help` or a dedicated `Examples:` block.)
   - Help text pairs human and `--output json` example invocations (no paired text + `--output json` example found within 5 lines in top-level or any subcommand `--help`. Pairing keeps agents from reverse-engineering the JSON invocation from the text one.)
   Requirements: https://anc.dev/p3

   fail
4. P4
   Fail-Fast, Actionable Errors

   stderr lacks a hint phrase. Looked for one of: try , use , expected , did you mean, see , run , for more, tip:, help:, hint:, suggestion:, available . Error messages must name what to do next, not just the symptom. · errors under `--output json` are not JSON-formatted. Consumers parsing stdout-as-JSON cannot recover the failure without a separate text-parsing path; switch the error writer to honor the active output mode.

   Remediation · P4⧉ copy prompt

   Make claude-code satisfy P4 (Fail-Fast, Actionable Errors) from the agent-native CLI standard. Address:
   - Error messages include a hint or remediation phrase (stderr lacks a hint phrase. Looked for one of: try , use , expected , did you mean, see , run , for more, tip:, help:, hint:, suggestion:, available . Error messages must name what to do next, not just the symptom.)
   - `--output json` produces JSON-formatted errors (errors under `--output json` are not JSON-formatted. Consumers parsing stdout-as-JSON cannot recover the failure without a separate text-parsing path; switch the error writer to honor the active output mode.)
   Requirements: https://anc.dev/p4

   fail
5. P5
   Safe Retries & Mutation Boundaries

   destructive subcommand(s) without `--force` or `--yes`: rm. Irreversible operations must require explicit confirmation so they can't be invoked accidentally. · write-pattern subcommand(s) present (purge, rm) but no read-pattern surface detected. If the CLI is write-only by design the MUST is satisfied vacuously; otherwise expose the read surface with agent-recognizable verbs (list/get/show/query/find/search).

   Remediation · P5⧉ copy prompt

   Make claude-code satisfy P5 (Safe Retries & Mutation Boundaries) from the agent-native CLI standard. Address:
   - Destructive subcommands require `--force` or `--yes` (destructive subcommand(s) without `--force` or `--yes`: rm. Irreversible operations must require explicit confirmation so they can't be invoked accidentally.)
   - Read and write surfaces are both visible in subcommand list (write-pattern subcommand(s) present (purge, rm) but no read-pattern surface detected. If the CLI is write-only by design the MUST is satisfied vacuously; otherwise expose the read surface with agent-recognizable verbs (list/get/show/query/find/search).)
   Requirements: https://anc.dev/p5

   fail
6. P6
   Composable, Predictable Command Structure

   pager referenced in --help but no --no-pager escape hatch advertised · 5/15 subcommand(s) follow standard verb names. Non-standard: agents, attach, auto-mode, gateway, import, mcp, purge, respawn, setup-token, ultrareview. MAY-tier — community-standard verbs (get/list/create/update/delete) help agents predict subcommand behavior across CLIs.

   Remediation · P6⧉ copy prompt

   Make claude-code satisfy P6 (Composable, Predictable Command Structure) from the agent-native CLI standard. Address:
   - Pager-using CLI ships --no-pager escape hatch (pager referenced in --help but no --no-pager escape hatch advertised)
   - Subcommand verbs follow community-standard names (5/15 subcommand(s) follow standard verb names. Non-standard: agents, attach, auto-mode, gateway, import, mcp, purge, respawn, setup-token, ultrareview. MAY-tier — community-standard verbs (get/list/create/update/delete) help agents predict subcommand behavior across CLIs.)
   - `--color` flag for explicit color control (no `--color` flag advertised. MAY-tier — `auto|always|never` lets agents and pipelines override the TTY-based default.)
   - Subcommand naming follows a consistent verb/noun convention (subcommand naming is inconsistent: 1 non-verb subcommand(s) (auth) mix verb and non-verb children at the second level, so an agent cannot predict where the action lives. SHOULD-tier: pick a consistent shape (all verb-first, all noun-verb hierarchy, or any combination where each non-verb group's children are uniformly verbs). The verb list is a heuristic; inspect `--help` to confirm.)
   Requirements: https://anc.dev/p6

   warn
7. P7
   Bounded, High-Signal Responses

   no --quiet/-q flag detected in --help output

   Remediation · P7⧉ copy prompt

   Make claude-code satisfy P7 (Bounded, High-Signal Responses) from the agent-native CLI standard. Address:
   - Quiet mode available (no --quiet/-q flag detected in --help output)
   Requirements: https://anc.dev/p7

   warn
8. P8
   Discoverable Through Agent Skill Bundles

   All 3 checks pass.

   pass