diff options
author | alex081512 <alex@rbbaader.de> | 2025-03-13 19:28:18 +0100 |
---|---|---|
committer | GitHub <noreply@github.com> | 2025-03-13 19:28:18 +0100 |
commit | 475912567ffc46e8bb288986310fbbf5052078df (patch) | |
tree | 01fbee5563ce695770735623b3c637248d0f1036 /src/plugins/lua/gpt.lua | |
parent | 33061ebe58aadd6047a269c695c9e74a34250c47 (diff) | |
download | rspamd-475912567ffc46e8bb288986310fbbf5052078df.tar.gz rspamd-475912567ffc46e8bb288986310fbbf5052078df.zip |
[Minor] Improve gpt prompt - Only ask for red flag when it's spam
Diffstat (limited to 'src/plugins/lua/gpt.lua')
-rw-r--r-- | src/plugins/lua/gpt.lua | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/src/plugins/lua/gpt.lua b/src/plugins/lua/gpt.lua index 98a3e38ee..9a20be551 100644 --- a/src/plugins/lua/gpt.lua +++ b/src/plugins/lua/gpt.lua @@ -958,14 +958,14 @@ if opts then "FROM and url domains. Evaluate spam probability (0-1). " .. "Output ONLY 3 lines:\n" .. "1. Numeric score (0.00-1.00)\n" .. - "2. One-sentence reason citing strongest red flag\n" .. + "2. One-sentence reason citing whether it is spam, the strongest red flag, or why it is ham\n" .. "3. Primary concern category if found from the list: " .. table.concat(lua_util.keys(categories_map), ', ') else settings.prompt = "Analyze this email strictly as a spam detector given the email message, subject, " .. "FROM and url domains. Evaluate spam probability (0-1). " .. "Output ONLY 2 lines:\n" .. "1. Numeric score (0.00-1.00)\n" .. - "2. One-sentence reason citing strongest red flag\n" + "2. One-sentence reason citing whether it is spam, the strongest red flag, or why it is ham\n" end end end |