me2resh/apexyard / .claude/skills/launch-check
launch-check
VerifiedProduction readiness audit — 10-dimension go/no-go sweep (security, a11y, compliance, analytics, SEO, GEO, perf, monitoring, docs, behaviour-quality).
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 29.2 KB |
| render-trend.shexecutable | script | 7.0 KB |
render-trend.sh
View on GitHub#!/bin/bash
# render-trend.sh — render the "Trend (last N runs)" section for /launch-check.
#
# Reads the runs directory (one JSON file per run, schema below), sorts by
# `ts`, takes the last N (default 5), and emits a markdown trend block:
# - heading
# - table (Date | Score | Verdict | Notes)
# - ASCII chart of scores over the same window
#
# Auto-notes column is derived from the score-delta vs the previous run
# (e.g. "Security +12, Perf +5"). The most recent row notes "This run".
# Operator-supplied notes are v2 (apexyard#183 v1 scope).
#
# Usage:
# render-trend.sh <runs_dir> [window]
#
# Behaviour:
# - Window defaults to 5.
# - With < 2 runs in the dir: prints nothing, exits 0. Callers (the
# /launch-check skill) suppress the trend section when there's no
# trend to show.
# - With >= 2 runs: prints the trend markdown block to stdout.
#
# JSON schema per run file (apexyard#183):
# {
# "ts": "2026-05-08T19:30:00Z", # ISO-8601 UTC, used for sort
# "branch": "main",
# "commit": "abc1234",
# "scores": {
# "security": 88, "accessibility": 94, "compliance": 76,
# "analytics": 90, "seo": 87, "performance": 68,
# "monitoring": 83, "docs": 91
# },
# "verdict": "go" | "go-with-warnings" | "conditional-go" | "no-go",
# "top_risks": ["...", "..."]
# }
#
# The headline score is the unweighted mean of scores.* — rounded to int.
# Forward-compatible: extra fields in the JSON are preserved as-is by
# /launch-check (the skill writes the file; this script only reads), so
# adopters with existing run files continue to work after framework upgrade.
#
# Dependencies: jq, sort, awk. No network calls.
set -u
RUNS_DIR="${1:-}"
WINDOW="${2:-5}"
if [ -z "$RUNS_DIR" ]; then
echo "usage: render-trend.sh <runs_dir> [window]" >&2
exit 2
fi
if [ ! -d "$RUNS_DIR" ]; then
# No runs dir => no trend. Treat as "first run".
exit 0
fi
if ! command -v jq >/dev/null 2>&1; then
echo "render-trend.sh: jq is required (install via brew install jq / apt install jq)" >&2
exit 2
fi
# ---------------------------------------------------------------------------
# Collect run files, sort chronologically by `ts`.
# Tab-separated: <ts>\t<path>. Sort by first column.
# ---------------------------------------------------------------------------
runs_tsv=$(
for f in "$RUNS_DIR"/*.json; do
[ -f "$f" ] || continue
ts=$(jq -r '.ts // ""' "$f" 2>/dev/null) || ts=""
[ -n "$ts" ] || continue
printf '%s\t%s\n' "$ts" "$f"
done | sort
)
run_count=$(printf '%s' "$runs_tsv" | grep -c . || true)
if [ "${run_count:-0}" -lt 2 ]; then
# Fewer than 2 runs => no trend section (matches AC: "if 2+ prior runs exist").
exit 0
fi
# Take the last $WINDOW lines.
window_tsv=$(printf '%s\n' "$runs_tsv" | tail -n "$WINDOW")
# ---------------------------------------------------------------------------
# Compute the headline score per run = round(mean(scores.*)).
# Build parallel arrays: dates[], scores[], verdicts[], score_blobs[].
# ---------------------------------------------------------------------------
dates=()
scores=()
verdicts=()
score_blobs=() # full scores object per run, for delta computation
while IFS=$'\t' read -r ts path; do
[ -n "$ts" ] || continue
date=$(printf '%s' "$ts" | cut -c1-10) # YYYY-MM-DD
verdict=$(jq -r '.verdict // "?"' "$path")
blob=$(jq -c '.scores // {}' "$path")
# Mean of the values in .scores. If empty, score = 0.
mean=$(jq -r '
(.scores // {}) as $s
| ($s | to_entries | map(.value) ) as $v
| if ($v | length) == 0 then 0
else (($v | add) / ($v | length)) | round
end
' "$path")
dates+=("$date")
scores+=("$mean")
verdicts+=("$verdict")
score_blobs+=("$blob")
done <<< "$window_tsv"
n=${#dates[@]}
if [ "$n" -lt 2 ]; then
# Defensive — shouldn't happen given the count check above.
exit 0
fi
# ---------------------------------------------------------------------------
# Per-row notes: diff this run's scores against the previous run's.
# Format: "Dim +N, Dim2 +M" for top 2 deltas by abs value. Negatives shown.
# Last row gets "This run" appended.
# ---------------------------------------------------------------------------
notes=()
for i in "${!dates[@]}"; do
if [ "$i" -eq 0 ]; then
notes+=("Initial baseline")
continue
fi
prev="${score_blobs[$((i-1))]}"
curr="${score_blobs[$i]}"
# jq computes deltas, sorts by abs desc, takes top 2, formats "Dim +N".
delta_str=$(jq -nr \
--argjson prev "$prev" --argjson curr "$curr" '
[ ($curr | to_entries[] )
| { dim: .key,
delta: (.value - ($prev[.key] // .value)) }
]
| map(select(.delta != 0))
| sort_by(-( .delta | if . < 0 then -. else . end ))
| .[0:2]
| map(
((.dim | .[0:1] | ascii_upcase) + (.dim | .[1:]))
+ " "
+ (if .delta > 0 then "+" else "" end)
+ (.delta | tostring)
)
| join(", ")
')
if [ -z "$delta_str" ]; then
delta_str="No change vs previous"
fi
if [ "$i" -eq $((n - 1)) ]; then
delta_str="${delta_str} (this run)"
fi
notes+=("$delta_str")
done
# ---------------------------------------------------------------------------
# Render the trend section.
# ---------------------------------------------------------------------------
echo "## Trend (last ${n} runs)"
echo ""
echo "| Date | Score | Verdict | Notes |"
echo "|------------|-------|------------------|-------|"
for i in "${!dates[@]}"; do
printf '| %-10s | %5s | %-16s | %s |\n' \
"${dates[$i]}" "${scores[$i]}" "${verdicts[$i]}" "${notes[$i]}"
done
echo ""
# ---------------------------------------------------------------------------
# ASCII chart — score vs time.
# Y-axis: range derived from min/max scores in window, padded by 5.
# X-axis: dates labeled below.
# ---------------------------------------------------------------------------
min=$(printf '%s\n' "${scores[@]}" | sort -n | head -1)
max=$(printf '%s\n' "${scores[@]}" | sort -n | tail -1)
# Pad and clamp to [0, 100].
y_lo=$(( min - 5 )); [ "$y_lo" -lt 0 ] && y_lo=0
y_hi=$(( max + 5 )); [ "$y_hi" -gt 100 ] && y_hi=100
[ "$y_hi" -le "$y_lo" ] && y_hi=$(( y_lo + 10 ))
# 5 horizontal grid lines, evenly spaced across [y_lo, y_hi].
rows=5
range=$(( y_hi - y_lo ))
echo "Score trend:"
echo ""
# Column width = 6 chars per run (matches MM-DD label width + 1 space).
COL_W=6
for ((r = rows - 1; r >= 0; r--)); do
level=$(( y_lo + (range * r) / (rows - 1) ))
line=$(printf '%3d |' "$level")
for j in "${!scores[@]}"; do
s="${scores[$j]}"
# Map s to a row index in [0, rows-1].
row_for_s=$(( ((s - y_lo) * (rows - 1) + range / 2) / range ))
if [ "$row_for_s" -eq "$r" ]; then
line="${line} ● " # 3 leading spaces + dot + 2 trailing = 6 cols
else
line="${line} "
fi
done
echo "$line"
done
# X axis line.
xaxis=" +"
for _ in "${!scores[@]}"; do xaxis="${xaxis}------"; done
echo "$xaxis"
# Date labels — abbreviated MM-DD, 6 chars each so labels don't collide.
labels=" "
for d in "${dates[@]}"; do
short=$(printf '%s' "$d" | cut -c6-) # MM-DD
labels="${labels}${short} "
done
echo "$labels"
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |