Auto-GPT-Plugins
Auto-GPT-Plugins copied to clipboard
Reduce "command returned too much output" errors in email plugin by truncating output
Motivation: As a continuation to the discussion with @riensen in PR https://github.com/Significant-Gravitas/Auto-GPT-Plugins/pull/177, this PR was created as a way to further reduce "command returned too much output" errors in the email plugin by truncating email output based on configuration settings.
Changes: Added truncate_emails function that truncates the email plugin's output based on the following requirements discussed in PR https://github.com/Significant-Gravitas/Auto-GPT-Plugins/pull/177.
We set three sensible limits:
MAX_CHARACTERS: The maximum number of characters for all emails retrieved
MAX_NUM_EMAILS: The maximum number of emails
MAX_EMAIL_CHARACTERS: The maximum length per body of each email
If MAX_CHARACTERS is breached, then we check how many emails are retrieved and set a return a smaller list
If the MAX_CHARACTERS is breached while MAX_NUM_EMAILS is not breched, then we start truncating each email which exceeds MAX_EMAIL_CHARACTERS starting with the longest one.
We also need to indicate that emails were truncated to Auto-GPT like adding
Testing: Added test to validate email truncation functionality
Future Work Consider whether to use ChatGPT API to summarize emails instead of truncating them.
Codecov Report
Patch and project coverage have no change.
Comparison is base (
808016e
) 58.52% compared to head (931553e
) 58.52%.
Additional details and impacted files
@@ Coverage Diff @@
## master #186 +/- ##
=======================================
Coverage 58.52% 58.52%
=======================================
Files 36 36
Lines 2122 2122
Branches 222 222
=======================================
Hits 1242 1242
Misses 858 858
Partials 22 22
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Do you have feedback about the report comment? Let us know in this issue.
Sorry I got assigned to this repo. Is this ready to go?
It is from my point of view, but I believe @riensen want additional tests added.