General-purpose coding policy for Baruch's AI agents
76
95%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
View guide
Low
Low-risk findings worth noting
#!/usr/bin/env bash
# Outcome-based tests for skills/herdr-teamlead/label-workspaces.sh.
#
# The fake herdr records every argv it is handed, so the assertions are about
# what was sent — including what was NOT sent on a re-run (rules/testing-
# standards.md; no live Herdr session).
#
# The harness drops `set -e` to aggregate results, so every fixture-setup
# command is checked explicitly and aborts with a fatal diagnostic on failure
# (rules/error-handling.md aggregate-reporting carve-out).
#
# Covers:
# 1. Labels the lot -> lead workspace, each worker's workspace and pane.
# 2. Pane label shape -> the kind alone; dispatch adds the role.
# 3. Idempotent -> a workspace already named is `unchanged`, unsent.
# 4. Overrides -> an explicit <agent>=<workspace-id> wins.
# 5. Caller excluded -> the lead's own pane is never relabelled as a worker.
# 6. Rename failure -> exit 3, the JSON names it, the run continues.
# 7. Roster failure -> exit 2.
# 8. Usage / env -> exit 1.
#
# Run: bash skills/herdr-teamlead/tests/test_label_workspaces.sh
set -uo pipefail
die() { echo "fatal: $*" >&2; exit 2; }
cleanup() { [[ -n "${TMP:-}" ]] && ! rm -rf "$TMP" && echo "warn: could not remove $TMP" >&2; return 0; }
pass() { PASS=$((PASS+1)); }
fail() { FAIL=$((FAIL+1)); echo " ✗ FAIL: $1" >&2; }
mk_fake_herdr() { # <path>
cat > "$1" <<'FAKE' || die "could not write the fake herdr"
#!/usr/bin/env bash
set -uo pipefail
[[ -n "${FAKE_ARGV_FILE:-}" ]] && printf '%s\n' "$*" >> "$FAKE_ARGV_FILE"
case "${1:-} ${2:-}" in
"agent list")
[[ -n "${FAKE_LIST_ERR:-}" ]] && { printf '{"error":{"code":"no_session"}}\n' >&2; exit 1; }
cat "${FAKE_LIST_FILE:?FAKE_LIST_FILE unset}"
exit 0
;;
"workspace list")
cat "${FAKE_WS_FILE:?FAKE_WS_FILE unset}"
exit 0
;;
"workspace rename"|"pane rename")
if [[ -n "${FAKE_RENAME_ERR:-}" && "$*" == *"${FAKE_RENAME_ERR}"* ]]; then
printf '{"error":{"code":"not_found"}}\n' >&2
exit 1
fi
printf '{"id":"cli:rename","result":{"ok":true}}\n'
exit 0
;;
esac
printf '{"error":{"code":"unsupported"}}\n' >&2
exit 2
FAKE
chmod +x "$1" || die "could not chmod the fake herdr"
}
run() { # [env...] -- then the script args come from ARGS
RUN_SEQ=$((RUN_SEQ+1))
ARGV="$TMP/argv.$RUN_SEQ"
: > "$ARGV" || die "could not create $ARGV"
OUT="$(env HERDR_ENV=1 HERDR_WORKSPACE_ID=w1 HERDR_PANE_ID=w1:p1 HERDR_BIN="$FAKE" \
FAKE_ARGV_FILE="$ARGV" FAKE_LIST_FILE="$LIST" FAKE_WS_FILE="$WS" "$@" \
bash "$SCRIPT" "${ARGS[@]}" 2>"$TMP/err.$RUN_SEQ")"
RC=$?
ERRTEXT="$(cat "$TMP/err.$RUN_SEQ")"
ARGVTEXT="$(cat "$ARGV")"
}
main() {
SCRIPT="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)/label-workspaces.sh"
[[ -f "$SCRIPT" && -r "$SCRIPT" ]] || die "label-workspaces.sh not found at $SCRIPT"
command -v jq >/dev/null 2>&1 || die "jq required for these tests"
TMP="$(mktemp -d -t label-ws-test.XXXXXX)" || die "mktemp failed"
trap cleanup EXIT
FAKE="$TMP/herdr"; mk_fake_herdr "$FAKE"
LIST="$TMP/list.json"
cat > "$LIST" <<'JSON' || die "could not write the roster"
{"id":"cli:agent:list","result":{"type":"agent_list","agents":[
{"agent":"claude","agent_status":"idle","pane_id":"w1:p1","workspace_id":"w1"},
{"agent":"codex","agent_status":"idle","pane_id":"w3:p1","workspace_id":"w3","name":"codex"},
{"agent":"grok","agent_status":"working","pane_id":"w4:p1","workspace_id":"w4","name":"grok"}
]}}
JSON
WS="$TMP/ws.json"
printf '{"id":"cli:workspace:list","result":{"workspaces":[{"workspace_id":"w1","name":""},{"workspace_id":"w3","name":""},{"workspace_id":"w4","name":""}]}}\n' > "$WS" \
|| die "could not write the workspace list"
FAIL=0; PASS=0; RUN_SEQ=0
local ARGS
# 1. Everything gets a name.
ARGS=(lead)
run
if [[ $RC -eq 0 ]] && printf '%s' "$OUT" | jq -e '
.lead.state == "renamed" and .lead.workspace_id == "w1"
and ((.agents | length) == 2)
and (.agents[0].workspace == "renamed") and (.agents[0].pane == "renamed")' >/dev/null 2>&1; then
pass; else fail "labels: got RC=$RC OUT=$OUT ERR=$ERRTEXT"; fi
if printf '%s' "$ARGVTEXT" | grep -q "workspace rename w1 lead" \
&& printf '%s' "$ARGVTEXT" | grep -q "workspace rename w3 codex"; then
pass; else fail "labels: expected the workspace renames, got ARGV=$ARGVTEXT"; fi
# 2. A pane starts as its kind; the workspace row above already names the
# agent, and the first dispatch relabels the pane with the role.
if printf '%s' "$ARGVTEXT" | grep -q "pane rename w3:p1 codex" \
&& printf '%s' "$ARGVTEXT" | grep -q "pane rename w4:p1 grok"; then
pass; else fail "pane label: expected the kind alone, got ARGV=$ARGVTEXT"; fi
# 5. The lead's own pane is a caller, not a worker (w1:p1 is excluded).
if ! printf '%s' "$ARGVTEXT" | grep -q "pane rename w1:p1"; then
pass; else fail "caller: the lead's pane must not be relabelled as a worker"; fi
# 3. A workspace already carrying its name is left alone.
printf '{"id":"cli:workspace:list","result":{"workspaces":[{"workspace_id":"w1","name":"lead"},{"workspace_id":"w3","name":"codex"},{"workspace_id":"w4","name":"grok"}]}}\n' > "$WS" \
|| die "could not rewrite the workspace list"
ARGS=(lead)
run
if [[ $RC -eq 0 ]] && printf '%s' "$OUT" | jq -e '
.lead.state == "unchanged" and (.agents | map(.workspace) | unique == ["unchanged"])' >/dev/null 2>&1; then
pass; else fail "idempotent: expected unchanged, got RC=$RC OUT=$OUT"; fi
if ! printf '%s' "$ARGVTEXT" | grep -q "workspace rename"; then
pass; else fail "idempotent: nothing may be sent for a name already in place"; fi
# 4. An explicit pair wins over the roster's workspace id.
ARGS=(lead codex=w9)
run
if [[ $RC -eq 0 ]] && printf '%s' "$ARGVTEXT" | grep -q "workspace rename w9 codex"; then
pass; else fail "override: expected w9, got ARGV=$ARGVTEXT"; fi
# 6. A label is cosmetic: a failure is reported, and the rest still run.
printf '{"id":"cli:workspace:list","result":{"workspaces":[{"workspace_id":"w1","name":""},{"workspace_id":"w3","name":""},{"workspace_id":"w4","name":""}]}}\n' > "$WS" \
|| die "could not rewrite the workspace list"
ARGS=(lead)
run FAKE_RENAME_ERR="workspace rename w3"
if [[ $RC -eq 3 ]] && printf '%s' "$OUT" | jq -e '
(.agents[] | select(.name == "codex") | .workspace) == "failed"
and ((.agents[] | select(.name == "grok") | .workspace) == "renamed")' >/dev/null 2>&1; then
pass; else fail "rename failure: expected exit 3 naming it, got RC=$RC OUT=$OUT"; fi
# 7. No roster, no labels.
ARGS=(lead)
run FAKE_LIST_ERR=1
if [[ $RC -eq 2 && -z "$OUT" ]]; then
pass; else fail "roster failure: expected exit 2, got RC=$RC OUT=$OUT"; fi
# 8. Usage and environment.
OUT="$(env HERDR_ENV=1 HERDR_WORKSPACE_ID=w1 HERDR_BIN="$FAKE" bash "$SCRIPT" 2>"$TMP/e8")"; RC=$?
if [[ $RC -eq 1 && -z "$OUT" ]] && grep -q "usage:" "$TMP/e8"; then
pass; else fail "usage: expected exit 1, got RC=$RC"; fi
OUT="$(env -u HERDR_ENV HERDR_BIN="$FAKE" bash "$SCRIPT" lead 2>"$TMP/e8b")"; RC=$?
if [[ $RC -eq 1 && -z "$OUT" ]] && grep -q "Herdr" "$TMP/e8b"; then
pass; else fail "outside Herdr: expected exit 1, got RC=$RC"; fi
# `env` stops option parsing at the first assignment, so -u comes first.
OUT="$(env -u HERDR_WORKSPACE_ID HERDR_ENV=1 HERDR_BIN="$FAKE" bash "$SCRIPT" lead 2>"$TMP/e8c")"; RC=$?
if [[ $RC -eq 1 && -z "$OUT" ]] && grep -q "HERDR_WORKSPACE_ID" "$TMP/e8c"; then
pass; else fail "no workspace id: expected exit 1, got RC=$RC"; fi
echo "─────────────────────────────────────────────" >&2
if [[ $FAIL -gt 0 ]]; then echo "FAILED: ${FAIL} failed, ${PASS} passed" >&2; exit 1; fi
echo "PASSED: all ${PASS} checks" >&2
}
if [[ "${BASH_SOURCE[0]}" == "$0" ]]; then
main "$@"
fi.tessl-plugin
hooks
rules
skills
adopt-fork-pr
herdr-standup
herdr-teamlead
references
teamlead
tests
migrate-to-plugin
onboard-repo
release
tests