Hua/nanobot
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix(slack): post-process slackify_markdown output to catch leftover artifacts

Browse Source 
The slackify_markdown library (markdown-it) fails to convert **bold** when
the closing ** is immediately followed by non-space text (e.g. **Status:**OK).
This is a very common LLM output pattern that results in raw ** showing up
in Slack messages.

Add _fixup_mrkdwn() post-processor that:
- Converts leftover **bold** → *bold* (Slack mrkdwn)
- Converts leftover ## headers → *bold* (safety net)
- Fixes over-escaped & in bare URLs
- Protects code fences and inline code from being mangled

Co-authored-by: Cursor <cursoragent@cursor.com>
This commit is contained in:
nanobot-agent
2026-02-24 12:44:17 +00:00
parent 8686f060d9
commit 81b669b36e
1 changed files with 10 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
nanobot/channels/slack.py
View File

@@ -230,7 +230,9 @@ class SlackChannel(BaseChannel):
_TABLE_RE = re.compile(r"(?m)^\|.*\|$(?:\n\|[\s:|-]*\|$)(?:\n\|.*\|$)*") _TABLE_RE = re.compile(r"(?m)^\|.*\|$(?:\n\|[\s:|-]*\|$)(?:\n\|.*\|$)*")
_CODE_FENCE_RE = re.compile(r"```[\s\S]*?```") _CODE_FENCE_RE = re.compile(r"```[\s\S]*?```")
_INLINE_CODE_RE = re.compile(r"`[^`]+`")
_LEFTOVER_BOLD_RE = re.compile(r"\*\*(.+?)\*\*") _LEFTOVER_BOLD_RE = re.compile(r"\*\*(.+?)\*\*")
_LEFTOVER_HEADER_RE = re.compile(r"^#{1,6}\s+(.+)$", re.MULTILINE)
_BARE_URL_RE = re.compile(r"(?<![|<])(https?://\S+)") _BARE_URL_RE = re.compile(r"(?<![|<])(https?://\S+)")
@classmethod @classmethod
@@ -247,8 +249,12 @@ class SlackChannel(BaseChannel):
def _fixup_mrkdwn(cls, text: str) -> str: def _fixup_mrkdwn(cls, text: str) -> str:
"""Fix markdown artifacts that slackify_markdown misses. """Fix markdown artifacts that slackify_markdown misses.
Handles: leftover ``**bold**``, ``&amp;`` in bare URLs, and The slackify_markdown library uses markdown-it which requires certain
collapsed paragraph spacing. boundary conditions for emphasis. Patterns like ``**key:**value``
(closing ``**`` immediately followed by non-space) are not recognised
as bold and pass through as literal asterisks. This method catches
those leftovers, stray ``## headers``, and over-escaped ``&amp;`` in
bare URLs, while leaving code spans untouched.
""" """
# Protect code blocks from further processing # Protect code blocks from further processing
code_blocks: list[str] = [] code_blocks: list[str] = []
@@ -258,9 +264,10 @@ class SlackChannel(BaseChannel):
return f"\x00CB{len(code_blocks) - 1}\x00" return f"\x00CB{len(code_blocks) - 1}\x00"
text = cls._CODE_FENCE_RE.sub(_save_code, text) text = cls._CODE_FENCE_RE.sub(_save_code, text)
text = cls._INLINE_CODE_RE.sub(_save_code, text)
# Fix leftover **bold** the library didn't convert (e.g. **key:**val)
text = cls._LEFTOVER_BOLD_RE.sub(r"*\1*", text) text = cls._LEFTOVER_BOLD_RE.sub(r"*\1*", text)
text = cls._LEFTOVER_HEADER_RE.sub(r"*\1*", text)
# Fix &amp; in bare URLs that the library over-escaped # Fix &amp; in bare URLs that the library over-escaped
text = cls._BARE_URL_RE.sub( text = cls._BARE_URL_RE.sub(