PostgreSQL error "invalid byte sequence for encoding UTF8" due to byte-wise truncation of commit_message

[petr-ciml]

Issue

Describe the bug
When Semaphore fetches a Git repository for a task, it stores the commit_message in the database. If the commit message exceeds the allowed column length, the Go backend truncates it. However, the truncation appears to be done using byte-wise slicing (e.g., msg[:200]) instead of rune-wise slicing.

If the truncation boundary falls exactly in the middle of a multi-byte UTF-8 character (e.g., Czech characters with diacritics like í, ž, á), it generates an invalid UTF-8 byte sequence. PostgreSQL strictly validates UTF-8 and immediately aborts the transaction.

This causes the Runner to silently fail with the UI logs showing:

Already up to date.
Get current commit hash
Get current commit message
Failed to update task commit
Failed to update TaskRunner status

And PostgreSQL logs throwing a critical error:

ERROR:  invalid byte sequence for encoding "UTF8": 0xc3
CONTEXT:  unnamed portal parameter $5
STATEMENT:  update task set status=$1, start=$2, "end"=$3, commit_hash=$4, commit_message=$5, runner_id=$6 where id=$7

To Reproduce

• Connect Semaphore to a PostgreSQL database.

• Push a Git commit where the message length hits the exact database truncation limit, and place a multi-byte UTF-8 character exactly on the boundary.

• Trigger the task.

• The task fails immediately when trying to update the task status in the DB.

Expected behavior
String truncation for database insertion should be UTF-8 aware to prevent database transaction abortions.

Environment:

• Semaphore UI Version: 2.17.33

• Database backend: PostgreSQL

Impact

Database

Installation method

Docker

Database

Postgres

Browser

Microsoft Edge

Semaphore Version

2.17.33

Ansible Version

Logs & errors

No response

Manual installation - system information

No response

Configuration

No response

Additional information

No response