teams-local-messages
Skill4 starsChanged 3 months agoMITView on GitHub
[PROTOTYPE] Read and recap Microsoft Teams conversations from the local Teams cache/database on macOS (new Teams client, WebView2 IndexedDB). Use when the user asks to access, search, list or summarise local Teams messages, channels or chats (e.g. a recap of the last week in Back-end and General) without using the Graph API.
- Installs packages
In this repository19 other files it gives its agents.
- AGENTS.mdAGENTS.md
- adapter-contract-testingSkill
- azure-devops-prSkill
- bugszero-root-causeSkill
- legacy-3pSkill
- port-refactoringSkill
- quality-gateSkill
- review-ai-writingSkill
- review-allSkill
- review-architectureSkill
- review-complexitySkill
- review-driftSkill
- review-impactSkill
- review-prSkill
- review-securitySkill
- review-slopSkill
- review-testingSkill
- scope-challengeSkill
- write-testsSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 50
Run topics to confirm the exact names the user means (a name like "Application" may not exist.
- Very specific: line 51
Run show for those topics with the requested --days.
- Distinctive: line 57
Only what Teams has cached locally: older history and unopened chats may be missing.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Teams local messages (macOS, new Teams)
Prototype. Built from a single exploratory session on one machine; the IndexedDB layout is an undocumented Teams internal and may change with any Teams update. Expect rough edges, and verify results against the Teams UI before relying on them.
Where the data is
New Teams (com.microsoft.teams2) keeps messages in a Chromium IndexedDB (LevelDB):
~/Library/Containers/com.microsoft.teams2/Data/Library/Application Support/Microsoft/MSTeams/
EBWebView/WV2Profile_tfw/IndexedDB/https_teams.microsoft.com_0.indexeddb.{leveldb,blob}Dead ends: Library/Caches/Microsoft/.../Cache is only HTTP cache; WebKit/WebsiteData/IndexedDB is empty; Cache.db has no messages.
Relevant databases (one per account, named Teams:<manager>:react-web-client:<id>:...):
conversation-manager/conversations: id → topic (channel or chat name)replychain-manager/replychains,replychains-2: records withmessageMapof messages (imDisplayName,originalArrivalTimein ms,contentas bytes of HTML,messageType,conversationId)
Setup (once)
pip install "git+https://github.com/cclgroupltd/ccl_chromium_reader.git" python-snappy(PyPI has no usable release.)
Usage
Script: scripts/teams_messages.py (copies the DB to a temp dir, so Teams can stay open and is never modified).
python3 scripts/teams_messages.py topics --days 7 # active conversations
python3 scripts/teams_messages.py show --topic back-end --topic general --days 7
python3 scripts/teams_messages.py export --out /tmp/teams/messages.jsonl--topic is a case-insensitive substring match on the conversation name.
Recap workflow
- Run
topicsto confirm the exact names the user means (a name like "Application" may not exist; say so and propose the closest ones). - Run
showfor those topics with the requested--days. - Summarise per conversation: decisions, PRs/work items (with ids), open questions, who asked what. Drop greetings and system noise.
- Mention gaps honestly (missing topics, only automatic meeting notices, etc.).
Caveats
- Only what Teams has cached locally: older history and unopened chats may be missing.
- 1:1/group chats often have no topic; they show as an empty topic or a thread id.
- Messages with no sender are system events (calls, pins, meeting notices); the script hides them.
- The cache is private data: keep exports in
/tmpor delete them afterwards; never share them externally. - macOS only; Windows uses a different path.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

