claudetools/errorlog.md

# Error Log

Brief records of preventable, pattern-worthy events across the fleet — used to improve
skills, write better CLAUDE.md rules, and clean stale/misleading memory. The aim: never
pay tokens twice for the same avoidable mistake. Append newest at the top; keep entries to
1-2 lines. **Always write via the helper, never by hand:**
`bash .claude/scripts/log-skill-error.sh "<skill/context>" "<brief>" [--correction|--friction] [--context "k=v"]`

Format: `YYYY-MM-DD | MACHINE | command/skill/context | [type] error (brief) [ctx: ...]`

Categories (the `[type]` tag): _(none)_ = skill/command execution failure ·
`[correction]` = user corrected an improper assumption I made ·
`[friction]` = preventable self-inflicted token-waste (harness/env/tool misuse; cite a
`ref=` in ctx when it repeats a documented gotcha — that flags a rule/memory to strengthen).

---

<!-- Append entries below this line -->

2026-06-18 | GURU-5070 | agy/search | gemini CLI threw ineligible/projectId setup error (throwIneligibleOrProjectIdError), empty response after 3 attempts [ctx: mode=search host=GURU-5070]

2026-06-18 | GURU-5070 | agy | gemini returned no response (empty after 3 attempts) [ctx: mode=search err=    at process.processTicksAndRejections (node:internal/process/task_queues:104:]

2026-06-18 | Howard-Home | git/sync-temp-files | [friction] controller-query scratch (.sta.json/.dev.json/.q*) written to repo CWD got swept into the commit by sync.sh 'git add -A', then a stray locked .sta.json blocked the rebase. Fix: write API scratch OUTSIDE the repo (or use the already-ignored .tmp- prefix); gitignored the patterns [ctx: ref=howard-home /tmp friction family]

2026-06-18 | Howard-Home | rmm | [friction] agent returns exit -1 'Failed to execute command' on a ~7KB multi-line powershell body sent as one command; split into <2KB section scripts and each ran fine [ctx: host=DESKTOP-TRCIEJA agent=0.6.66]

2026-06-18 | GURU-5070 | coord/ad2-comms | [correction] tried to coordinate with the AD2 session via coord API msg+lock; AD2 is network-isolated (Gitea only, no coord API) so those were no-ops. ALL inter-session comms with AD2 must go via git /sync (committed notes/docs).

2026-06-18 | GURU-5070 | syncro | comment POST piped straight to jq failed with 'jq: parse error: Invalid numeric literal at line 1 col 10' and left it AMBIGUOUS whether the note posted (GET-verify showed it had NOT); per no-retry rule had to GET first, then re-post. Robust pattern that worked: jq -n payload to a file, POST with --data-binary @file, capture response to a file, then GET-verify by subject. Skill's curl|jq comment pattern should adopt this. [ctx: ticket=32441 skill=syncro pattern=curl-pipe-jq]

2026-06-18 | GURU-5070 | post-bot-alert | Discord POST failed (non-200/unreachable) [ctx: channel=#bot-alerts http=400 resp={"message": "The request body contains invalid JSON.", "code": 50109}]

2026-06-18 | GURU-5070 | ssh/ad2 | [correction] attributed AD2 SSH timeouts to a flaky VPN tunnel + my rapid scp/ssh bursts; real cause = OpenVPN adapter MTU 1500 vs tunnel PMTU ~1424 -> TCP MSS blackhole that drops bulk/scp segments (DF set) while small cmds pass. Fix: tunnel adapter MTU 1400 [ctx: ref=feedback_prefer_ssh_over_rmm]

2026-06-18 | GURU-5070 | bash/env | [friction] /tmp curl-write then Windows-python read mismatch; wrote .claude/tmp + absolute path fixed it [ctx: ref=feedback_tmp_path_windows]

2026-06-18 | Howard-Home | pfsense-ssh/logs | [friction] used clog on pfSense 25.07 logs (now plain-text ASCII) -> empty output -> wrongly concluded DHCP log was empty / dhcpd not serving; cost a hypothesis. Read pfSense 25.07 logs with tail/grep/cat directly, NOT clog [ctx: ref=reference_pfsense_25_07_ops client=cascades-tucson]

2026-06-18 | AD2 | vault | real vault.sh not found at resolved vault_path; vault read failed [ctx: path=D:/vault/scripts/vault.sh]

2026-06-18 | AD2 | vault | real vault.sh not found at resolved vault_path; vault read failed [ctx: path=D:/vault/scripts/vault.sh]

2026-06-17 | GURU-5070 | mailbox/365-mail | [correction] claimed in a prior session that /mailbox skill + memories were repointed off the deleted fabb3421 to the 365-mail suite, but mailbox.md still hardwired fabb3421 (token 401 AADSTS700016). Correct app is the dedicated ComputerGuru Mailbox app 1873b1b0 via get-token.sh 'mailbox' tier (cert auth); repointed mailbox.md + feedback_365_remediation_tool.md 2026-06-17. Lesson: verify the edit actually landed before reporting it done.

2026-06-17 | Howard-Home | wiki-compile/coord | [friction] skill doc Phase 6 shows 'lock release claudetools wiki/<type>/<slug>' but coord.py takes 'lock release <id>'; wasted a round-trip. Capture the lock id from claim output and release by id. [ctx: ref=wiki-compile-skill]

2026-06-17 | Howard-Home | unifi/controller-write | [friction] UniFi OS controller PUT (rest/device port_overrides) returned 403 without CSRF. Fix: login with -D headers, read 'x-updated-csrf-token' (or decode csrfToken from TOKEN cookie JWT), send as X-CSRF-Token on PUT/POST/DELETE

2026-06-17 | Howard-Home | bash/env | [friction] Git-Bash /tmp path mismatch again: msys curl -o /tmp/x.json wrote where Windows python could not read it (FileNotFoundError). Fix: write API JSON to CWD-relative ./.x.json so curl+python share the path [ctx: ref=howard-home known /tmp friction]

2026-06-17 | Howard-Home | pfsense/cascades-voice-vlan | [correction] assumed new RFC1918 alias + DNS-to-firewall:53/123 rules + clone VLAN20 for VOICE isolation; correct is clone the GUEST VLAN (VLAN50/igc1.50, the only actually-isolated net: 3x literal-CIDR quick blocks + pass any) and hand out PUBLIC DNS 8.8.8.8/1.1.1.1 via DHCP. VLAN20 is NOT isolated; config.xml rules were mismapped/not matching live pfctl -sr [ctx: ref=voice-vlan-cutover.md; lesson=read pfctl -sr not just config.xml]

2026-06-17 | GURU-5070 | ssh/plink-windows | [friction] plink -pw with a special-char password through Git-bash -> native plink.exe got MANGLED (CommandLineToArgvW) -> 'Access denied', led to a wrong 'stale cred' conclusion + wasted DMs. The pw was correct. Use system OpenSSH KEY auth (or pass pw via stdin/file), never plink -pw, for special-char passwords on Windows [ctx: ref=feedback_windows_quote_stripping;host=192.168.0.9]

2026-06-17 | GURU-5070 | vault/d2testnas-cred | [correction] CORRECTION: the d2testnas password is NOT stale - it worked for Mike. My plink '-pw' failure was special-char mangling through Git-bash -> native plink.exe (Windows CommandLineToArgvW). Use KEY auth (system OpenSSH) for password-special-char hosts [ctx: ref=feedback_windows_quote_stripping;host=192.168.0.9]

2026-06-17 | GURU-5070 | vault/d2testnas-cred | vaulted clients/dataforth/d2testnas credentials.password REJECTED by 192.168.0.9 (Access denied) - stale/wrong; needs update. Key auth is the reliable path [ctx: host=192.168.0.9]

2026-06-17 | GURU-5070 | agy | gemini returned no response (empty after 3 attempts) [ctx: mode=search err=Attempt 1 failed: You have exhausted your capacity on this model. Your quota wil]

2026-06-17 | GURU-5070 | agy | gemini auth/login failure [ctx: mode=search]

2026-06-17 | Howard-Home | wiki-compile | [friction] Phase 6 release cmd documented as 'coord.py lock release claudetools <resource-path>' but coord.py 'lock release' takes the LOCK ID, not the resource path -> inline release no-ops and strands the lock until TTL. Fix: capture lock id from claim and release by id. [ctx: skill=wiki-compile phase=6]

2026-06-17 | GURU-5070 | grok | grok xsearch incomplete (rc=124); auto-fell back to gemini [ctx: mode=xsearch]

2026-06-17 | GURU-5070 | grok | grok xsearch returned no result [ctx: mode=xsearch stopReason=]

2026-06-17 | GURU-5070 | research-method | [correction] treated blind endpoint-probing as 'authoritative' over web search; Mike: web searches (grok/gemini) have been MORE valuable - they gave the real leads (connector proxy, teleport setting path), probing only confirmed and mostly 404s. Lead with web search; probe only to CONFIRM a search/doc-derived hypothesis [ctx: ref=feedback_interview_ai_read_docs]

2026-06-17 | GURU-5070 | bash/background-ai | [friction] mixing a backgrounded ask-grok/ask-gemini ('&' + wait) with foreground curl probes in ONE Bash command repeatedly yields an EMPTY output capture; run AI calls as separate run_in_background Bash tool calls, never '&'+wait inline with work to capture [ctx: ref=grok/gemini wrappers]

2026-06-17 | GURU-5070 | agy | gemini returned no response (empty after retry) [ctx: mode=search err=Ripgrep is not available. Falling back to GrepTool.]

2026-06-17 | GURU-5070 | grok | grok xsearch returned no result [ctx: mode=xsearch stopReason=]

2026-06-17 | GURU-5070 | grok | grok xsearch returned no result [ctx: mode=xsearch stopReason=]

2026-06-17 | GURU-5070 | grok | grok xsearch returned no result [ctx: mode=xsearch stopReason=]

2026-06-17 | Howard-Home | unifi-wifi/apply-radio | [friction] per-AP --apply loop: each call re-logs-in to the controller; rapid succession throttles -> write silently skips (no [ok]). Fix: space calls (sleep 3-4) or add multi-AP/one-login support

2026-06-17 | Howard-Home | wiki-compile | [friction] full recompile Sonnet subagent ran ~54min then crashed on 32k output cap (tried to emit the ~490-line article despite being told to write-to-file and return only a summary); recovered via direct surgical Edits to the existing article. Fix: for --full on large existing articles, prefer targeted Edit integration over a subagent rewrite, or hard-cap/forbid article body in the subagent reply. [ctx: skill=wiki-compile target=client:cascades-tucson]

2026-06-17 | GURU-5070 | grok/ask-grok.sh | [correction] blindly probed grok with slow timed runs to find the xsearch syntax instead of reading its bundled docs (~/.grok/docs/user-guide/ + README.md) / interviewing the model first; the docs gave the root cause directly (web_search=grok-4.20-multi-agent multi-agent model, killed by --no-subagents) [ctx: ref=feedback_interview_ai_read_docs]

2026-06-17 | GURU-5070 | grok | grok xsearch returned no result [ctx: mode=xsearch stopReason=]

2026-06-17 | GURU-5070 | grok | grok xsearch returned no result [ctx: mode=xsearch stopReason=]

2026-06-17 | GURU-5070 | grok/xsearch | xsearch returned empty (stopReason finalization quirk) on Grok-co-work research; fell back to text-mode design + Claude synthesis [ctx: topic=grok-as-claude-cowork]

2026-06-17 | GURU-5070 | grok | grok xsearch returned no result [ctx: mode=xsearch stopReason=]

2026-06-17 | GURU-5070 | syncro/customer-comment | [friction] posted a customer-facing emailed comment to #32333 WITHOUT previewing for human review first; preview-before-send is mandatory for ALL outgoing comms [ctx: ref=CLAUDE.md syncro 'Before any POST: Always show the full payload and wait for confirmation']

2026-06-17 | GURU-5070 | syncro/customer-comment | [correction] conflated Arizona Medical Transit (AMT-PC, Windows: Dell bloatware + misbehaving Syncro agent) cleanup into the Scileppi Mac (#32333) customer comment; Scileppi was a full-disk/Trash + Mail + Downloads-redesign job, no Dell/agent work [ctx: ticket=32333 mac=scileppi]

2026-06-16 | GURU-5070 | grok | grok xsearch returned no result [ctx: mode=xsearch stopReason=]

2026-06-16 | GURU-5070 | rmm/shell-quoting | [friction] shutdown /r /t 60 /c "comment" FAILED via command_type=shell: embedded double-quotes in the command got mangled through the agent cmd layer, shutdown rejected the args and dumped usage. Fix: avoid embedded double-quotes in RMM shell commands (drop /c, or build the quoted arg another way) [ctx: ref=bash/curl.exe-on-windows quote-stripping; agent=AMT-PC; cmd=shutdown]

2026-06-16 | GURU-5070 | syncro/billing | [friction] billed AMT as non-prepaid off the customer-SEARCH endpoint prepay=null; real prepay (detail endpoint) was 7.0 -> invoice correctly netted $0 via block, but I set the wrong (upsell) invoice note. Always read prepay from /customers/{id} detail, not the list/search [ctx: ref=syncro_invoice_verification_pattern;cust=7088349]

2026-06-16 | GURU-5070 | rmm/onboarding-diagnostic | onboarding-diagnostic.ps1 fails on Win7/PowerShell 2.0: uses [ordered] hashtables (PS3+) -> 'Unable to find type [ordered]', empty DIAG-JSON, no grade. Probe not PS2/legacy-compatible -> can't diagnose Win7/2008R2 legacy agents (first hit: AMT-PC)

2026-06-16 | GURU-5070 | rmm-diagnose | could not extract valid diagnostic JSON from probe output [ctx: host=AMT-PC status=completed exit=0]

2026-06-16 | GURU-5070 | mailprotector/starrpass | [correction] assumed starrpass.com was on Mailprotector; correct: starrpass.com is direct-to-MS (EOP/Defender) - Starr Pass MP account 16170 covers ONLY devconllc.com. Check @starrpass.com mail via remediation-tool/EOP

2026-06-16 | GURU-5070 | mailprotector | HTTP 404 POST https://emailservice.io/api/v1/users/find_by_address: "Not found" [ctx: cmd=find-user]

2026-06-16 | Howard-Home | bash/curl.exe-on-windows | [friction] PowerShell-invoked curl.exe strips embedded double-quotes from --data-urlencode args (CommandLineToArgvW), silently mangling POST bodies; pfSense PHP became 'echo PHPRUNS-OK' -> 'Undefined constant'. Fix: write payloads with single-quotes only, build $ via [char]36, keep one line. [ctx: ref=pfsense diag_command.php php-exec; cost=4 wasted RMM round-trips]

2026-06-16 | GURU-5070 | remediation-tool/get-token | [friction] get-token.sh reads vault_path from ~/.claude/identity.json (home), which lacks the field on this machine; repo identity.json (.claude/identity.json) has it. Fix: export VAULT_ROOT_ENV=$(jq -r .vault_path .claude/identity.json) before calling get-token [ctx: ref=remediation-tool;machine=GURU-5070]

2026-06-15 | GURU-5070 | rmm/quickbooks-folderbrowser | [correction] assumed F:FolderRedirection was a dead/missing drive (Test-Path F: = False under SYSTEM); correct: F: is a per-user NETWORK-mapped redirected folder, invisible to the SYSTEM context RMM runs in - must diagnose mapped-drive/redirect issues in user_session

2026-06-15 | GURU-5070 | rmm | ProfWiz Pro silent-install command returned 'Execution error: Failed to execute command' (status failed, no stdout) on SP-SharonW11 [ctx: agent=86de13d7 host=SP-SharonW11 task=upw-install]

2026-06-15 | GURU-5070 | remediation-tool (Starr Pass licensing) | [correction] reported Brian Shinn's account as DELETED (tied it to the recycle-bin bshinn@ from 6/10 onboarding); actually Mike UNLICENSED Brian in M365 - account not deleted. Don't conflate a soft-deleted recycle-bin entry with the user's recent action; check the ACTIVE account's assignedLicenses for an unlicense

2026-06-15 | GURU-5070 | grok | grok xsearch returned no result [ctx: mode=xsearch stopReason=]

2026-06-15 | GURU-5070 | grok | grok xsearch returned no result [ctx: mode=xsearch stopReason=Cancelled]

2026-06-15 | GURU-5070 | graduation-pipeline (BEAST Ollama) | [friction] BEAST Ollama ran inference on CPU (api/ps showed qwen3:32b AND qwen3:14b with vram=0); 32b timed out at 240s, 14b at 175s. GPU not engaged - the 'use BEAST GPU' premise needs a BEAST-side Ollama GPU config/driver fix before large-model triage is practical

2026-06-15 | GURU-5070 | graduation-pipeline (BEAST env) | [friction] assumed BEAST uses WSL because 'bash' there resolved to the WindowsApps WSL stub (uname said WSL2). BEAST runs the harness under Git-for-Windows MSYS bash like other Windows boxes; reach its Ollama via localhost (Git-bash) or the Tailscale IP. REPEAT of the documented WSL-stub-vs-Git-bash gotcha [ctx: ref=feedback_windows_bash_mapping]

2026-06-15 | GURU-5070 | tmp-promotion-check (/save,/scc) | [friction] hung for minutes: line 51 ran 'grep -rqlF <f> projects/' per scratch file, recursing Rust target/, node_modules/, .git in the guru-rmm/guru-connect submodules. Fixed: --include='*.md' + --exclude-dir for heavy trees. Stalled the /save sync behind it

2026-06-15 | GURU-5070 | memory-dream (--apply-safe) | flagged feedback_broken_backlinks_are_writeme_markers.md as an orphan and appended a DUPLICATE index line though it already had one — orphan detector likely keys on the frontmatter name: slug, not the (file.md) link target. Fix the index-line matching to compare by filename [ctx: mode=apply-safe]

2026-06-15 | GURU-5070 | powershell/var-case | [friction] PowerShell vars are case-INSENSITIVE: $gUid silently overwrote $guid (GPO id), Set-ADObject hit a bad DN and left GPT.ini/AD versionNumber inconsistent until fixed. Never rely on case to distinguish PS variables

2026-06-15 | GURU-5070 | python/argv-limit | [friction] passed full /api/agents JSON (248 agents) as a python CLI arg -> 'Argument list too long' on Windows. Pipe large payloads via stdin, not argv

2026-06-15 | GURU-5070 | bash/env-persist | [friction] re-derived RMM token every call after $TOKEN/$RMM vanished between Bash tool calls - shell env does NOT persist across calls; must re-eval auth (or chain) in the same command

2026-06-15 | GURU-5070 | bash/tmp-path | [friction] wrote curl -o /tmp/x.json then jq read it back and failed (No such file) - Git-Bash vs Write/tool /tmp resolve differently. Pipe directly or use repo-relative paths. REPEAT of documented gotcha [ctx: ref=feedback_tmp_path_windows]

2026-06-15 | GURU-5070 | DMARC / DNS | [correction] assumed ACG's own INKY rua convention (reports-sg.inkydmarc.com) applied to a client domain; only use the INKY rua if THAT client is onboarded to INKY - otherwise plain p=none or a real mailbox

2026-06-15 | GURU-5070 | remediation-tool (sendMail) | [correction] assumed none of the consented apps could send mail and started granting Graph Mail.Send; the Exchange Operator app ALREADY had Graph Mail.Send - I was decoding the EXO-audience token, not a Graph-audience token. Mint a Graph token for the app before concluding a permission is missing

2026-06-15 | GURU-5070 | rmm-search | [correction] assumed the CLI search must replicate the UI Omnibox scoreMatch exactly; user wants a FLEXIBLE forgiving multi-field search optimized for first-try correctness, not UI parity


2026-06-15 | GURU-BEAST-ROG | /syncro (comment edit) | Syncro API does not expose a comment-edit or comment-delete endpoint — once posted, comments can only be modified via the GUI. Bot posted an internal resolution note with an unwanted "Performed by: ClaudeTools Discord Bot" line and could not remove it programmatically. Remediation needed: either suppress bot-attribution lines from internal notes by default, or add a GUI-edit step to the workflow when the note needs correction.

2026-06-14 | GURU-5070 | mailbox skill (Graph token) | FABB app `fabb3421` (Claude-MSP-Access / "Cloud MSP Access") token request returned AADSTS700016 — app/SP no longer present in azcomputerguru.com tenant (deleted; gotchas.md already marked it deprecated). Blocks /mailbox + the M365 contacts task. Verified the remediation suite (live, ACG tenant) carries NO Mail.Send/Mail.ReadWrite/Contacts scopes (investigator has Mail.Read only) — so a straight repoint can't restore mailbox-send/contacts. Pending Mike decision: stand up a single-tenant ACG-internal mailbox app vs. add scopes to a suite tier. [2026-06-15] Docs hardened — gotchas.md now marks fabb3421 DELETED with the Mail/Contacts-scope blast radius + flags the 3 legacy "old app only" tenants (Valleywide/Dataforth/Cascades) as now having NO working remediation app (migration URGENT); mailbox.md carries a BLOCKED/AADSTS700016 banner. DECISION 2026-06-15 (Mike): Mail.Send goes into the suite (Exchange Operator tier) since its real use is IR victim-notification during mailbox takeovers; add Mail.Send to the exchange-op manifest + consent, repoint mailbox.md to exchange-op. Implementation not yet executed (production app change, needs go).

2026-06-14 | GURU-KALI | coord skill (coord.py) | Documented invocation `py .claude/skills/coord/scripts/coord.py ...` failed exit 127 — `py` (the Windows py-launcher) does not exist on Linux. Worked around with `python3`. [RESOLVED 2026-06-14] Added `.claude/scripts/py.sh` (resolves the working interpreter: identity.json `python.command` -> py -> python3 -> python, skipping the MS Store shim) and repointed all skill/command DOC invocations from bare `py` to `bash "$CLAUDETOOLS_ROOT/.claude/scripts/py.sh"`. The `.sh` skill scripts already resolved internally — left untouched. Broadcast to fleet.

2026-06-14 | GURU-BEAST-ROG | coord skill (coord.py msg send) | `py "$CLAUDETOOLS_ROOT/.claude/skills/coord/scripts/coord.py"` failed — `$CLAUDETOOLS_ROOT` is not exported in fresh Git-bash shells here, so the path resolved under `C:\Program Files\Git\`. [RESOLVED 2026-06-14] Added `.claude/scripts/ensure-settings-env.py` (seeds `env.CLAUDETOOLS_ROOT` in per-machine `settings.local.json` from `identity.json`); Claude Code injects it into every Bash call. Wired into ONBOARDING.md + broadcast to fleet. Effective next session start.

2026-06-14 | GURU-BEAST-ROG | /sync (sync.sh Phase 3, submodule update) | submodule `projects/msp-tools/guru-rmm` checkout of f38da05 aborted: untracked `docs/RMM_THOUGHTS.md` would be overwritten. Parent repo synced fine; submodule pointer left lagging. Recurring transient. [RESOLVED 2026-06-15] sync.sh now has `resolve_submodule_collisions()` — on the abort it moves only the untracked files the incoming commit tracks aside to `<file>.synced-aside-<UTCstamp>` (content preserved, NOT --force) then retries once. Verified live: guru-rmm advanced ed92097->f38da05; the aside copy held 94 lines of un-committed 2026-06-08 thoughts (rescued, not lost — needs manual merge into canonical RMM_THOUGHTS.md).
2026-06-16 | HOWARD-HOME | unifi-wifi/device-control.sh provision | cmd/devmgr force-provision returned HTTP 400 (mac 0c:ea:14:3f:40:6d / AP 445); verb needs fix — likely wrong cmd name or requires device _id not mac. block/kick/locate via stamgr work; adopt/restart/upgrade unverified.