Choose your rules of engagement

Your pace. Your risk. Your call. What Safety on, Weapons tight and Weapons free change in your day, and which one to pick.

Your rules of engagement are one choice for how hard and how fast Blowhorn pushes on a platform. There are three:

The names are the military's weapons-control states, from most to least restrictive; the plain words beside each say what it does, so you never need the military meaning to choose well. One choice sets pace, pointer movement, caps, spacing, quiet hours, outreach, the pause after a sign-in challenge, retries and how much runs at once, together. The headline is ownership-first - "Your pace. Your risk. Your call." The app shows the new names; the CLI and the config file keep the stored values (blowhorn risk set cautious, defaults.risk_profile: cautious, --risk-profile cautious), and the reference still calls the setting the Risk Profile, so old scripts keep working.

Choose it in the app on the Posture screen: the slider at the top is the choice for everyone, a track from Safety on (slow end, in the brand's teal blue) through Weapons tight (teal) to Weapons free (fast end, in the warn colour). The same choice shows as a mini-slider in the menu-bar popover, which opens Posture, and changing it anywhere changes it everywhere. Or use blowhorn risk set cautious. blowhorn risk show says which preset is in effect and where it came from, and blowhorn risk grid lists every profile on every platform. Any single dial you set yourself wins over the preset for that dial, and the rest of the preset still applies around it; a profile with a dial set outside the preset is a custom profile, not a fourth preset you can pick.

Set one platform for everyone

To run one platform more carefully for every profile, for example LinkedIn on Safety on while the rest stay on Weapons tight:

1. Open Posture.
2. In the grid, pick the chip in the Each platform row under LinkedIn, and choose Safety on.

Or, from Terminal:

$ blowhorn risk set cautious --platform linkedin

The choice is kept with your organization, so every computer you run Blowhorn on uses it, and the row's caption says every computer. If your organization cannot keep it yet, it is saved on this computer only: the command says scope: this_computer with the reason, and the row says this computer only. Your organization's choice for that platform comes first, so when it already has one, the command says it still applies on every computer and this computer's choice has no effect until an organization admin clears it. Choose Inherit in the same chip, or run blowhorn risk set default --platform linkedin, to clear it.

Set one profile, or one profile on one platform

• One profile on every platform: pick the chip in that profile's Every platform column, or blowhorn risk set cautious --profile company.
• One profile on one platform: pick that profile's cell under the platform, or blowhorn risk set brisk --profile company --platform x.

A chip you chose is filled; a dashed chip is inherited, and the line under it says where it comes from ("from LinkedIn", "from this profile", "from everyone"). To put a cell back, point at it and press its reset, or choose Inherit in its menu; the menu names what it would inherit first.

Which choice wins

The more specific choice wins: one activity beats one profile on one platform, which beats a platform's or a profile's choice, which beats everyone's. When a platform and a profile both have a choice, the more careful one wins, so "LinkedIn on Safety on for everyone" also slows a profile set to Weapons free, and "X on Weapons free for everyone" never speeds up a profile set to Safety on. To let one profile go faster on one platform, choose its cell. A new profile or computer still runs under Safety on for its first 14 days. The full order is in Which choice wins.

Adjust the limits yourself

Under the grid, The limits Blowhorn keeps starts from the preset: change the pace and daily caps per platform in the by platform table, and the other limits below it, including typing and pointer timing, then save limits. The header reads "Custom, based on Weapons tight, 3 dials changed" once dials move, and reset to preset puts every dial back. No dial may exceed a built-in ceiling: your own cap can only lower a ceiling, never raise it. Two dials carry a driver note: pauses and mistakes run on the Playwright driver only. The extension driver, which types through the Blowhorn extension in Google Chrome, sends one key per message and runs no pause plan, so on that driver it types with the shaped delay from the delay band and no pauses or corrections. The full values are in the configuration reference, with a stable anchor per dial in Risk Profile dials.

Safety on (cautious): one thing at a time, nights off

Pick Safety on for a new account, a new computer, or an unattended run you will not be watching.

• Between actions: at least 15 minutes, plus a random wait of up to 15 minutes, between two public actions on one platform for one profile. A day of posting is therefore a slow trickle, not a burst.
• How much per day: a quarter of each platform's built-in ceiling. On LinkedIn that is 2 posts and 10 comments a day (before any cap you set yourself); on X it is 12 posts a day. Your own daily caps can only take these lower.
• Nights: nothing public from 22:00 to 07:00 local time. A due activity inside those hours waits until morning; it is not failed.
• After a challenge: if a platform shows a CAPTCHA, checkpoint or restriction page, that profile on that platform is held for 72 hours from the last such record.
• Outreach: off entirely - no connects, invites or follows.
• At once: at most 1 run, and one profile at a time on a platform. A second run is refused and names your cautious profile; its activities stay queued.
• Retries: a failed scheduled job is not retried.

Weapons tight (standard): half the ceiling, with the guardrails on

Weapons tight is the built-in default, and the right choice when an account is established and you want Blowhorn at a normal pace, held to half of each platform's built-in ceiling.

• Between actions: no minimum spacing; pace still shapes typing cadence and the random pauses between actions.
• How much per day: half of each platform's built-in ceiling. On LinkedIn that is 5 posts and 20 comments a day; on X it is 25 posts a day, before any cap you set yourself or a warmup ramp.
• Nights: no preset quiet hours. A quiet window you name yourself with defaults.quiet_hours still holds scheduled claims.
• After a challenge: the profile on that platform is held for 24 hours.
• Outreach: on for established platforms, off on platforms still being built (Facebook, Instagram, TikTok, YouTube).
• At once: one profile at a time on a platform, and no preset limit on runs overall. Setting defaults.concurrency.max_runs adds one.
• Retries: a failed scheduled job is not retried.

Weapons free (brisk): fast, while you watch

Pick Weapons free for an established account you are watching, when you want the most throughput and accept the most exposure.

• Between actions: no minimum spacing, and the fast pace.
• How much per day: the full built-in ceiling - Blowhorn's own estimate of a safe day, not a number the platform publishes. On LinkedIn that is 10 posts and 40 comments a day; on X it is 50 posts a day. Weapons free never raises a ceiling, and never goes past a limit the platform publishes; it only lifts the share the other two keep below it (a quarter, a half), along with Safety on's quiet hours, spacing and outreach limits.
• Nights: no preset quiet hours.
• After a challenge: the profile on that platform is held for 12 hours - the shortest pause, so a challenge is the place Weapons free costs you the most caution.
• Outreach: on everywhere, including platforms still being built.
• At once: different profiles may share a platform, and no preset limit on runs overall. One profile is still held to one run on a platform.
• Retries: a failed scheduled job gets one retry. A login is never retried, and a retry is never granted for an unknown outcome.

A new profile warms up under Safety on, whatever you chose

A new profile or a new Mac runs under Safety on for scheduled jobs until it has history - 14 days, whatever preset you picked. Its caps ramp over those 14 days, from about a quarter of Safety on's caps on day one (about 1 LinkedIn post a day) to Safety on's full caps by day 14, and only then does your chosen preset take over. A profile idle longer than the 15-day ledger window warms up again. A run you start yourself is not held by the warmup, only warned about, and it follows your chosen preset's caps, quiet hours and outreach.

If you are still unsure

• New account, or you will not watch it: Safety on.
• Established account, normal use: Weapons tight.
• Established account, you are watching, and you want the most done today: Weapons free.
• One account needs to be slower than the rest: set that profile or platform to Safety on, and leave the rest alone - the choice is per profile and per platform, not only computer-wide.
• One platform feels risky for everyone: set that platform to Safety on in the Each platform row.

One activity on its own rules of engagement

One recurring activity can run under Safety on while everything else on the same profile and platform runs under Weapons tight: open the activity, set rules of engagement (inherit means the activity's own choice is blank, exactly as before), and save. From the terminal the same value is risk_profile=cautious in the activity's params. The activity's detail then shows the preset and where it came from (Safety on · this activity, or Safety on · inherits: LinkedIn × work profile), and while the profile is warming it also says what the tick enforces (warming: runs Safety on for now). An activity without one profile and one platform names no single preset; its targets resolve their own. For a single run you start by hand, blowhorn --risk-profile cautious post ... overrides everything for that run only.

An activity's own choice stays as you set it, even when it is looser than the Posture page: an activity on Weapons free runs Weapons free while the same profile is on Safety on everywhere else. (A choice on one order in New order works the other way and can only tighten.) Two things still stand above it: the 14-day warm-up, and a limit the platform publishes.

A limit the platform publishes is the one hard stop

Where a platform publishes its own limit for an account, nothing in Blowhorn goes past it: not an activity's own choice, not Weapons free, not your own dials, not a setting in .blowhorn.yaml or the environment, not --ignore-risk and not do it now. Blowhorn cuts any cap that would go past one down to it, and says so: the run prints which setting asked for more and the platform's page it read the limit from, and the routine editor says this routine stops at … under rules of engagement when the preset you pick would go past it. When the limit is already reached for the day, the activity waits until it clears.

Only limits the platform itself publishes count, each with its page and the day it was read:

The rest of the numbers on this page are Blowhorn's own estimates, and there is no hard stop where a platform publishes nothing: LinkedIn, Reddit, Hacker News, Slack and GitHub publish no per-account count for what Blowhorn does. Platforms change these limits without notice, and Blowhorn reads them again on a schedule.

Related

• Configuration reference - every dial, every preset value.
• Risk Profile dials - one anchor per dial, for the desktop info buttons.
• You're in control - the rules of engagement among every safety control.
• How Blowhorn paces itself - measured pacing and the no-retry rule.