Implements issue #654 Option A (embedded-sample convention) plus the
verification harness the issue asked for:
- scripts/smoke_json_output.py — new advisory gate (G9) that discovers
every tool whose --help advertises JSON output, runs <tool> --sample
<json-flag>, and asserts the stdout parses as JSON. Tools advertising
JSON without --sample are reported as 'uncovered' (a backlog, not a
failure) so the gate can be adopted incrementally; --strict flips that
to a hard failure once coverage is high. Wired into ci-quality-gate.yml
alongside G8.
- Added --sample embedded fixtures to the 5 tools named in #654:
error_budget_calculator, slo_review, blast_radius_calculator,
audit_log_analyzer, api_linter. Their required args are now optional
when --sample is passed; missing-arg behavior is unchanged otherwise.
- Fixed 4 tools the new gate surfaced (prompt_rater, coach_tip_classifier,
cheat_code_filter, redaction_linter): their --sample path printed human
text and ignored --json; it now honors the JSON flag.
- Synced the 3 dual-published standalone copies (slo-architect x2,
chaos-engineering) so the drift guard stays green.
Gate now reports 16 tools covered, 16 verified, 0 failures.
https://claude.ai/code/session_01CUWsrUNZP9jpxvAwq67UiT
CI's structure validator and Tessl quality review both expect the
repo's canonical Path-B layout (plugin root with .claude-plugin/,
agents/, commands/, skills/<name>/ subfolders) used by every other
engineering skill — caveman, grill-me, handoff, write-a-skill.
The flat layout shipped initially scored 42.9/100 internally and 0/100
on Tessl. After wrapping:
- SKILL.md and references moved under skills/claude-coach/
- 3 stdlib Python tools added: cheat_code_filter, prompt_rater,
coach_tip_classifier (5-gate decision tree)
- cs-claude-coach agent persona + /cs:claude-coach slash command
- .claude-plugin/plugin.json manifest (canonical "./skills/claude-coach" form)
- SKILL.md frontmatter extended with Tier/Category/Author/Dependencies
fields the validator requires; spec body content preserved
- SKILL.md picks up Name/Description/Features/Usage/Examples sections
appended below the original body so the validator's required-sections
check passes without disturbing the original coaching instructions
- coach_tip_classifier inlines a minimal prompt scorer instead of
cross-importing prompt_rater, eliminating the auditor's sys.path
finding
Local verification:
- Structure validator: 100.0/100 EXCELLENT (0 failed checks)
- Security auditor: PASS (0 critical/high/info findings)
- All 3 scripts pass --help and --sample
https://claude.ai/code/session_01KUWQ32LyRMFxYPN4VnhJsY
A coaching layer that teaches Claude power-user techniques on first
activation, then surfaces at most one tip per response when it spots a
missed opportunity. Hard rules: answer the user first, max one tip per
turn, silence is the default. Ships SKILL.md, a ranked cheat-codes
glossary, and explicit coaching rules to prevent over-coaching.
https://claude.ai/code/session_01KUWQ32LyRMFxYPN4VnhJsY
