One Reddit post can turn into 420 comments, 69 reports, and three mods opening the same thread. Firewatch turns that into a shared incident room: one queue item, clear reasons it needs attention, one mod claiming ownership, and a visible action log.
Firewatch is built for posts that need more than a single approve/remove click: scam domains, heated arguments, support-safety cleanup, repeated wording, report spikes, AutoModerator filters, and threads where the mod team needs shared context before acting.
Firewatch is an open source project released under the MIT License—built for the community to use, fork, and enhance to fit each subreddit’s needs. Repository: github.com/takitajwar17/firewatch.
- Creates a subreddit-level Firewatch board post that acts as the shared mod queue.
- Adds posts to the queue when they’re sent manually from a post menu item, or when triggers detect signals like reports, AutoModerator filters, watched words, watched domains, comment activity, and mod actions.
- Shows a simple 0–5 attention rating and plain-English reasons (so a mod can triage fast).
- Uses a claim workflow so the team can see who’s handling an incident (and avoid duplicate work).
- Keeps Reddit-native moderation actions close to the evidence (approve/remove/spam, lock/unlock, ignore reports, crowd control, flair, user actions like mute/ban, and more—permission gated).
- Supports configurable Automations / Response Rules that can suggest actions, prepare them for approval, or run selected actions when explicitly enabled.
- Includes demo drills to seed realistic incidents for playtesting in a sandbox subreddit.
- Removes stored post/comment content when Reddit sends delete triggers.
- Install Firewatch in a subreddit you moderate.
- Open the subreddit menu and choose Open Firewatch.
- Open Firewatch settings to tune watched words, watched domains, rating thresholds, reminder text, signal weights, and available mod actions.
- Let Firewatch queue matching activity from triggers, or choose Send to Firewatch from a post menu when a thread needs attention.
- Review the Firewatch rating, visible reasons, impact snapshot, comments, users, activity, mod notes, and matched automations.
- Claim the post, approve acceptable comments, remove rule-breaking comments, lock the thread, run prepared automation actions, or save a handoff note.
- Mark the incident resolved once the queue item no longer needs active review.
Firewatch Automations watch posts, comments, reports, incident ratings, and user strike counts. An automation can:
- Suggest actions without running them.
- Prepare actions for a moderator to review and run.
- Auto-run safe Firewatch-only actions while leaving Reddit-native actions for moderator approval.
- Auto-run all selected actions when a mod team deliberately enables that mode for a trusted rule. Reddit-native auto-run actions still require the post to be claimed, and Firewatch records the claimant as the responsible actor.
Automations are transparent: each match records why it matched, which actions were prepared, which actions ran, and which actions were skipped. Moderators should test automations in a sandbox subreddit before enabling automatic actions in an active community.
Firewatch ratings are advisory. The app explains its reasons, but moderators remain responsible for reviewing context, applying community rules, and deciding whether to approve, remove, lock, ban, mute, flair, ignore reports, or take any other action. Firewatch is designed to reduce queue work, not replace moderator judgment.
Firewatch is a moderator-only app. All menu items are registered with
forUserType: "moderator", and the server also checks the current moderator's
subreddit permissions before returning mod data or accepting actions.
- To open the review queue, a moderator must be allowed to manage posts and
comments. Reddit calls this
posts, orallfor full mod access. - To change Firewatch settings, automations, demo data, or reset app data, a
moderator must be allowed to change subreddit settings. Reddit calls this
config, orallfor full mod access. - To ban, mute, approve users, add mod notes, clear Firewatch strikes, or clean
up a user's recent content, a moderator must be allowed to manage users.
Reddit calls this
access, orallfor full mod access. User-content cleanup also needs post and comment moderation access. - To set or clear post flair, a moderator must be allowed to manage both posts
and post flair. Reddit calls these
postsandflair, orallfor full mod access. - Moderators who cannot change subreddit settings do not receive watched lists, rating thresholds, automation rules, or automation logs in the webview payload. They can still see the review data, available action controls, and reminder text needed for post moderation.
- Moderators who cannot manage post flair do not receive post flair templates or run flair actions.
If someone without enough mod access opens Firewatch, the app shows a plain access screen instead of loading private queue, settings, automation, or action data.
Firewatch stores only the data needed to run the moderation workflow for installed communities: incident state, public Reddit identifiers, public content excerpts needed for review, moderator settings, automation logs, action history, user strike summaries, and handoff notes. Incident, claim, and automation log records expire after 30 days. Per-moderator selected incident state expires after 24 hours. Community configuration, automations, user strike summaries, and the Firewatch queue post reference do not have an app-set expiry and remain until changed, cleared, reset by app logic, or removed by platform storage behavior.
Firewatch is a deterministic signal queue, not a complete abuse detector. It handles common watched-word, watched-domain, report, repeated wording, and reply-cluster patterns, including simple domain obfuscation, but moderators should still review context and tune community-specific rules.
See Privacy Policy and Terms of Service for the full policy text.