fix(privacy): skip PII redaction on Discord/Slack (mentions need real IDs)

Discord uses <@user_id> for mentions and Slack uses <@U12345> — the LLM
needs the real ID to tag users. Redaction now only applies to WhatsApp,
Signal, and Telegram where IDs are pure routing metadata.

Add 4 platform-specific tests covering Discord, WhatsApp, Signal, Slack.
This commit is contained in:
teknium1
2026-03-16 05:58:34 -07:00
parent c51e7b4af7
commit 9a423c3487
3 changed files with 42 additions and 2 deletions

View File

@@ -183,6 +183,16 @@ class SessionContext:
}
_PII_SAFE_PLATFORMS = frozenset({
Platform.WHATSAPP,
Platform.SIGNAL,
Platform.TELEGRAM,
})
"""Platforms where user IDs can be safely redacted (no in-message mention system
that requires raw IDs). Discord is excluded because mentions use ``<@user_id>``
and the LLM needs the real ID to tag users."""
def build_session_context_prompt(
context: SessionContext,
*,
@@ -196,10 +206,14 @@ def build_session_context_prompt(
- What platforms are connected
- Where it can deliver scheduled task outputs
When *redact_pii* is True, phone numbers are stripped and user/chat IDs
When *redact_pii* is True **and** the source platform is in
``_PII_SAFE_PLATFORMS``, phone numbers are stripped and user/chat IDs
are replaced with deterministic hashes before being sent to the LLM.
Platforms like Discord are excluded because mentions need real IDs.
Routing still uses the original values (they stay in SessionSource).
"""
# Only apply redaction on platforms where IDs aren't needed for mentions
redact_pii = redact_pii and context.source.platform in _PII_SAFE_PLATFORMS
lines = [
"## Current Session Context",
"",

View File

@@ -130,3 +130,27 @@ class TestBuildSessionContextPromptRedaction:
p1 = build_session_context_prompt(ctx1, redact_pii=True)
p2 = build_session_context_prompt(ctx2, redact_pii=True)
assert p1 != p2
def test_discord_ids_not_redacted_even_with_flag(self):
"""Discord needs real IDs for <@user_id> mentions."""
ctx = _make_context(user_id="123456789", platform=Platform.DISCORD)
prompt = build_session_context_prompt(ctx, redact_pii=True)
assert "123456789" in prompt
def test_whatsapp_ids_redacted(self):
ctx = _make_context(user_id="+15551234567", platform=Platform.WHATSAPP)
prompt = build_session_context_prompt(ctx, redact_pii=True)
assert "+15551234567" not in prompt
assert "user_" in prompt
def test_signal_ids_redacted(self):
ctx = _make_context(user_id="+15551234567", platform=Platform.SIGNAL)
prompt = build_session_context_prompt(ctx, redact_pii=True)
assert "+15551234567" not in prompt
assert "user_" in prompt
def test_slack_ids_not_redacted(self):
"""Slack may need IDs for mentions too."""
ctx = _make_context(user_id="U12345ABC", platform=Platform.SLACK)
prompt = build_session_context_prompt(ctx, redact_pii=True)
assert "U12345ABC" in prompt

View File

@@ -839,7 +839,7 @@ privacy:
redact_pii: false # Strip PII from LLM context (gateway only)
```
When `redact_pii` is `true`, the gateway redacts personally identifiable information from the system prompt before sending it to the LLM:
When `redact_pii` is `true`, the gateway redacts personally identifiable information from the system prompt before sending it to the LLM on supported platforms:
| Field | Treatment |
|-------|-----------|
@@ -849,6 +849,8 @@ When `redact_pii` is `true`, the gateway redacts personally identifiable informa
| Home channel IDs | Numeric portion hashed |
| User names / usernames | **Not affected** (user-chosen, publicly visible) |
**Platform support:** Redaction applies to WhatsApp, Signal, and Telegram. Discord and Slack are excluded because their mention systems (`<@user_id>`) require the real ID in the LLM context.
Hashes are deterministic — the same user always maps to the same hash, so the model can still distinguish between users in group chats. Routing and delivery use the original values internally.
## Speech-to-Text (STT)