harness: fleet-wide functional-error + correction + friction logging

Add .claude/scripts/log-skill-error.sh — the canonical agent error log helper
(writes errorlog.md in DATE | MACHINE | skill | [type] error format, soft-fails).
Three categories: execution failures (default), user corrections (--correction),
and preventable self-inflicted friction (--friction; cite ref= when it repeats a
documented gotcha). Goal: stop paying tokens twice for the same avoidable mistake.

- CLAUDE.md: make logging mandatory for all skills + corrections + friction.
- skill-creator: new skills must wire in the helper (guidance + checklist).
- Retrofit every skill script's genuine failure branches to call the helper
  (b2/bitdefender/mailprotector/packetdial/coord python CLIs; remediation-tool
  + onboard365 bash; vault, rmm-auth, post-bot-alert, agy, grok, 1password,
  run-onboarding-diagnostic). Handled conditions + self-tests left alone.
- errorlog.md: broaden header to cover skills + harness + corrections; seed this
  session's corrections (INKY, Mail.Send token-audience, omnibox-strictness) and
  friction (git-bash /tmp, env-persistence, argv-limit, PowerShell var-case).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
2026-06-15 11:39:43 -07:00
parent 10128b3669
commit da87d314c5
29 changed files with 388 additions and 36 deletions

View File

@@ -239,6 +239,7 @@ case "$AUTH_OVERRIDE" in
if [[ -z "$CERT_X5T" || -z "$CERT_KEY_B64" ]]; then
echo "ERROR: REMEDIATION_AUTH=cert but cert fields missing in vault ($VAULT_PATH)" >&2
echo " Required fields under credentials: cert_thumbprint_b64url, cert_private_key_pem_b64" >&2
bash "$CLAUDETOOLS_ROOT/.claude/scripts/log-skill-error.sh" "remediation-tool" "get-token: cert auth forced but cert fields missing in vault" --context "tier=$TIER vault=$VAULT_PATH" >/dev/null 2>&1 || true
exit 4
fi
AUTH_METHOD="cert"
@@ -251,6 +252,7 @@ case "$AUTH_OVERRIDE" in
if [[ -z "$CLIENT_SECRET" ]]; then
echo "ERROR: REMEDIATION_AUTH=secret but client_secret missing in vault ($VAULT_PATH)" >&2
echo " Check field: credentials.client_secret (or credentials.credential for older entries)" >&2
bash "$CLAUDETOOLS_ROOT/.claude/scripts/log-skill-error.sh" "remediation-tool" "get-token: secret auth forced but client_secret missing in vault" --context "tier=$TIER vault=$VAULT_PATH" >/dev/null 2>&1 || true
exit 4
fi
AUTH_METHOD="secret"
@@ -269,6 +271,7 @@ case "$AUTH_OVERRIDE" in
echo "ERROR: no usable credential found in $VAULT_PATH" >&2
echo " Need either credentials.cert_thumbprint_b64url + credentials.cert_private_key_pem_b64," >&2
echo " or credentials.client_secret (or legacy credentials.credential)." >&2
bash "$CLAUDETOOLS_ROOT/.claude/scripts/log-skill-error.sh" "remediation-tool" "get-token: no usable credential (cert or client_secret) in vault" --context "tier=$TIER vault=$VAULT_PATH" >/dev/null 2>&1 || true
exit 4
fi
AUTH_METHOD="secret"
@@ -336,6 +339,7 @@ PY
if [[ $ASSERT_RC -ne 0 || -z "$CLIENT_ASSERTION" ]]; then
echo "ERROR: failed to build client_assertion JWT" >&2
[[ -n "$CLIENT_ASSERTION" ]] && echo "$CLIENT_ASSERTION" >&2
bash "$CLAUDETOOLS_ROOT/.claude/scripts/log-skill-error.sh" "remediation-tool" "get-token: failed to build client_assertion JWT (cert auth)" --context "tier=$TIER rc=$ASSERT_RC" >/dev/null 2>&1 || true
exit 4
fi
@@ -371,11 +375,13 @@ if [[ -z "$TOKEN" ]]; then
echo " After the admin accepts, run onboard-tenant.sh to assign required directory roles:" >&2
SCRIPT_DIR_ERR="$(dirname "${BASH_SOURCE[0]}")"
echo " bash ${SCRIPT_DIR_ERR}/onboard-tenant.sh ${TARGET}" >&2
bash "$CLAUDETOOLS_ROOT/.claude/scripts/log-skill-error.sh" "remediation-tool" "get-token: AADSTS7000229 — app not consented in tenant" --context "tenant=$TENANT_ID tier=$TIER auth=$AUTH_METHOD" >/dev/null 2>&1 || true
exit 5
fi
echo "ERROR: token request failed (tenant=$TENANT_ID tier=$TIER auth=$AUTH_METHOD)" >&2
echo "$RESP" >&2
bash "$CLAUDETOOLS_ROOT/.claude/scripts/log-skill-error.sh" "remediation-tool" "get-token: token request failed (no access_token)" --context "tenant=$TENANT_ID tier=$TIER auth=$AUTH_METHOD err=${ERROR_CODE:-none}" >/dev/null 2>&1 || true
exit 5
fi