compute_age
ChatGPTCompute a bag's age in days off roast and its resting/staling verdict — the coffee's current bag by default, or the bag named.
Other
Overview
Dial in better espresso, one informed adjustment at a time. Set up your beans, grinder, espresso machine, and scale once. After each shot, log the dose, yield, brew time, and how it tasted. ShotPulled considers the full context—including the coffee’s age and the way your grinder is calibrated—then recommends the single most useful adjustment for your next shot and explains why. ShotPulled also keeps your coffee inventory accurate automatically. Every logged dose is deducted from the bag, so you can see the grams remaining and an estimate of how many shots are left based on the dose you actually use. It tracks the bag’s peak window too, showing how many peak days remain and the date that window ends. When a recipe works, lock it in and return to it anytime. ShotPulled supports grinder settings across clicks, numbered collars, compound ring positions such as 1.2.2, and named settings, with built-in specifications for 67 grinders, 124 espresso machines, and 24 scales. Your ShotPulled account stays in sync across ChatGPT, the web dashboard, and the iPhone app. Register a bag here, and it’s already waiting on your phone when you’re ready to pull your next shot.
Tools
delete_shot
ChatGPTDelete a logged shot from history. Restores the bean's remaining weight (which is derived from logged doses). Hard delete — there is no undo. To fix a mistake on an otherwise-valid shot, prefer update_shot over delete-and-relog.
diagnose_preview
ChatGPTSandbox-only diagnosis, writes nothing (no verdict, no recommendation-trail entry). Two modes: pass shot_id to dry-run a LOGGED shot (optionally overriding its sensory_tags — the "what would this read as?" preview; the shot's own bean and its age at pulled_at are used), or pass the full metric set (bean_id, grinder_id, machine_id, grind_label, dose_g, yield_g, time_s, source) for a hypothetical shot. Identical output shape to diagnose_shot, including bean_context.
diagnose_shot
ChatGPTEvaluate shot metrics to provide ranked hypotheses (C6) and warn on flip-flops (G2) or fatigue (G6). Resolves the SHOT's own bean — age computed at the shot's pulled_at — never the active profile, and echoes it as bean_context, so diagnosing an older or differently-filed shot is always safe. The engine reads metrics and sensory tags — NOT free-text notes — so make sure taste feedback is recorded as sensory_tags on the shot (via log_shot or update_shot) before diagnosing; otherwise an in-range shot that tastes bad will come back "balanced".
get_dial_state
ChatGPTRetrieve current dialing state, active context, recent shots, recommendations, and locked recipes.
get_kb_version
ChatGPTRetrieve the current version of the knowledge base.
get_maintenance_guides
ChatGPTStep-by-step maintenance guides — tools, supplies, steps and warnings, in our own words with the maker's documents as sources. With no arguments, every guide for the user's current grinders and machines (and the water filter fitted in each). With guide_ids, those guides; the ids come from maintenance.tasks[].guide.id in list_grinders and list_machines. Ids that match no guide are listed under missing.
get_rule
ChatGPTRetrieve the detailed text of a specific rule from the knowledge base by its rule ID.
get_stats
ChatGPTGet compact flat usage stats for a grinder, machine, or bean.
grinder_math
ChatGPTDetermine steps/clicks difference and directions between setting labels (C3).
kb_changelog
ChatGPTRetrieve the changelog showing updates and version differences for the knowledge base.
list_beans
ChatGPTList the coffees registered for the account, with status filter. Each coffee carries its bags[] (every purchase, with its own freshness and grams left) and current_bag_id; the row's top-level roast_date, state, grams_remaining and age fields are the current bag's. The
id on each row is the number this account knows that coffee by, counting from 1 — safe to show, and what other tools expect back; bag ids are a separate series.list_grinders
ChatGPTList grinders registered for the account, with status filter. The
id on each row is the number this account knows that record by, counting from 1 — safe to show, and what other tools expect back. Each row carries its maintenance block: every task with its class (S schedule, H habit, Y symptom, A the machine signals it, T technician, N never, U unstated, E event), due state, and the guide to follow.list_machines
ChatGPTList machines registered for the account, with status filter. The
id on each row is the number this account knows that record by, counting from 1 — safe to show, and what other tools expect back. Each row carries its maintenance block, researched per model: every task with its class, due state, what the maker forbids, and the guide to follow. settings lists the selectors its model has and where each stands (null: unknown model, or none).list_maintenance_due
ChatGPTWhat maintenance is due across the user's current grinders and machines, most urgent first — for "is anything due?". Only tasks with an interval compete: a task the machine counts and signals itself, one done by habit or when a symptom shows, and one the maker advises against never appear. Each item says how far along it is (ratio, done of every, on which basis), whether it is coming up (soon), and whether the interval is the user's, the maker's or ShotPulled's default.
list_recipes
ChatGPTList recipes (all or filtered by equipment and status). The
id on each row is the number this account knows that record by, counting from 1 — safe to show, and what other tools expect back.list_scales
ChatGPTList registered scales for the account. The
id on each row is the number this account knows that record by, counting from 1 — safe to show, and what other tools expect back.list_shots
ChatGPTList shot history with support for pagination and filtering. Every shot carries a derived taste_pending flag (1 = logged with no sensory tags, no rating, and no tasted flip — the tasting is still owed and can be backfilled with update_shot).
list_waters
ChatGPTList registered waters for the account. The
id on each row is the number this account knows that record by, counting from 1 — safe to show, and what other tools expect back.lock_recipe
ChatGPTLock a successful shot as the reference dialing recipe for this equipment profile. It overwrites the recipe already locked for the same coffee, grinder, machine, program and drink intent, which keeps its recipe_id.
log_maintenance
ChatGPTLog a cleaning or maintenance job against a piece of equipment. A task that takes the thing apart opens a calibration epoch, and every recipe locked before it is read through that epoch's offset — so record the offset here if you counted the clicks while reassembling, which is the one moment anybody actually knows it. Any task in the maintenance vocabulary can be logged against a grinder or machine it belongs to — including one the maker advises against, because the log records what was done. Read each piece of equipment's tasks, their intervals and their due state from list_grinders and list_machines.
log_recommendation
ChatGPTLog a recommended dialing change to allow contradiction/oscillation checking (G2).
log_shot
ChatGPTLog an espresso or alternative shot/brew attempt. Uses the active context (bean/grinder/machine/program) unless overridden via bean_id/grinder_id/machine_id/program_id — ALWAYS pass bean_id explicitly when the user names a specific coffee, so the shot cannot land on the wrong bag. Backdate with pulled_at when the shot happened earlier. When the user reports taste (bitter, sour, harsh, hollow…), include matching sensory_tags — the diagnosis engine reads tags and metrics, not free-text flavor_notes. When the user tasted the shot and reports it as GOOD/clean (no defects), pass tasted=1 with no sensory_tags — otherwise a tag-less, rating-less shot is filed taste-pending as if never tasted.
merge_beans
ChatGPTFold one coffee into another when the same coffee was registered twice. Every bag, shot, recipe and recommendation of from_bean_id moves to into_bean_id; the active coffee follows if it was the one merged away; origins move only if the target has none. Where both hold a locked recipe for the same grinder, machine, program and drink, the newer lock wins and the older is deleted. The source coffee is archived, never deleted. Irreversible — confirm both numbers with the user first (list_beans shows them).
register_bag
ChatGPTRegister another bag of a coffee that is already in list_beans — the same roaster's product, bought again. It keeps the coffee's recipes, notes and history; only the purchase facts (roast date, weight, price, state) are new. Does not change the active coffee. Returns a bag_id: bag numbers are their own series, separate from bean numbers. For a coffee that is not in list_beans yet, call register_coffee instead.
register_coffee
ChatGPTRegister a coffee that is NOT yet in list_beans, together with its first bag. It does not switch a coffee in use: it becomes active only when no coffee is in use (none active, or the active one archived or with every bag finished); otherwise call set_active with its bean_id when the user starts brewing it. The
active field of the reply says which. The coffee is the roaster's product (roaster, origins, variety, process, roast level, roasting type, cupping score, aromatics, roaster notes, rest window); the bag is this purchase (roast date, weight, price, state). For another bag of a coffee already on file call register_bag instead — it keeps the recipes and notes. Returns bean_id and the first bag's bag_id.register_grinder
ChatGPTRegister a manual or electric grinder with its dial settings scheme. It becomes the active grinder only when no grinder is in use (none active, or the active one archived); otherwise call set_active with its grinder_id to switch to it. The
active field of the reply says which.register_machine
ChatGPTRegister an espresso or alternative preparation machine. It becomes the active machine only when no machine is in use (none active, or the active one archived); otherwise call set_active with its machine_id to switch to it. The
active field of the reply says which.register_program
ChatGPTRegister a programmed shot button on a machine. A program cuts the shot off either by weight (program_type "volumetric", set target_volume_g) or by time (program_type "timed", set target_time_s). One machine can have a mix of both. A program_number the machine already has is overwritten with these settings; shots logged on it keep pointing at it.
register_scale
ChatGPTRegister a brewing scale — the instrument that weighs the dose and the yield.
register_water
ChatGPTRegister a custom or commercial water formulation.
set_active
ChatGPTSwitch the active bean, grinder or machine for this account. The numbers come from list_beans, list_grinders and list_machines. Registering a coffee, grinder or machine makes it active only when nothing of that kind is in use (the
active field of the reply says which), so call this to start using something new as well as to switch back.set_equipment_archived
ChatGPTArchive or restore a grinder, machine, water, or scale.
set_grinder_position
ChatGPTSet the current grinder collar position. source="measured" resets verification freshness. It replaces the stored position; no history of earlier positions is kept.
set_maintenance_offset
ChatGPTRecord how far a maintenance job moved the reference, on an entry that opened a calibration epoch. Takes an ENTRY, not a piece of equipment: the offset belongs to the event that caused it, which is what lets several of them compose for a recipe locked before all of them. Zero is a legitimate measurement and means "measured, and it had not moved" — omit offset_value entirely to clear it back to unmeasured.
set_maintenance_schedule
ChatGPTSet the user's own schedule for one maintenance task on one grinder or machine: their own interval, switching it off, or muting the reminder until a date. Only the keys you send in fields change; send every: null to go back to the maker's guidance and muted_until: null to unmute. Use it when the user states how often they do a task or asks to be left alone about one — never to invent an interval. The task's class, interval and due state are read back from list_grinders and list_machines.
suggest_next_step
ChatGPTRetrieve the single next experiment step from the reasoning engine.
update_bag
ChatGPTChange bag-level facts on one bag of a coffee: open, seal, freeze, thaw or finish it, correct its roast date, weight, grams left or price. Bag numbers come from the bags[] on list_beans (or current_bag_id). For the coffee itself — roaster, origin, roast level, rest window, notes — use update_bean.
update_bean
ChatGPTUpdate a coffee's own fields (roaster, origins, roast level, rest window, notes, rating…). Bag-shaped fields sent here (state, roast_date, bag_weight_g, grams_remaining, cost…) land on the coffee's CURRENT bag — use update_bag when it has more than one. Changing dial_category changes what the engine advises (G5, C10), so say why in the optional note.
update_grinder
ChatGPTCorrect a registered grinder — its name, burrs, collar scheme, nominal step, layout, range or notes. Only the keys you send in fields change; send null to clear an optional field (an empty string also clears a text field, but not a number or a closed list). Required fields can be changed but not cleared. fields may be omitted when only reason is sent. setting_scheme, nominal_step and components_spec describe the collar every recorded grind label is read through: call with dry_run first to see which recorded settings would read as a different position, and tell the user before saving. Saving records the change in the grinder's maintenance history; when a locked recipe's setting moves it also opens an unmeasured calibration epoch (set_maintenance_offset resolves it). The collar's current setting is marked assumed when it reads differently, or cleared when the new collar cannot read it, and measured offsets are converted when a click changes worth. Changing direction_convention is a correction — it moves no number, but every finer/coarser said about this grinder was the wrong way round.
update_machine
ChatGPTCorrect a registered machine — its name, control type, prep type and style, basket, boiler, accessories or notes — or record where its selectors stand (settings). Only the keys you send in fields change; send null to clear an optional field (an empty string also clears a text field, but not a number or a closed list). Required fields can be changed but not cleared. fields may be omitted when only reason is sent. prep_style decides whether shots on this machine are diagnosed at all: only "espresso" is. Programs are edited with register_program, the built-in grinder link is not editable here, and archiving is set_equipment_archived.
update_scale
ChatGPTCorrect a registered scale — its type, resolution, capacity, capabilities, build or notes. Only the keys you send in fields change; send null to clear an optional field (an empty string also clears a text field, but not a number or a closed list). Required fields can be changed but not cleared. fields may be omitted when only reason is sent. The three capabilities are 1, 0, or null for "the maker never said"; leave one out rather than sending 0 when nobody knows.
update_shot
ChatGPTCorrect fields on an already-logged shot in place — no need to delete and re-log. Use for fixing a wrong dose/yield/time or grind label, re-filing a shot onto the right bean (bean_id), backfilling rating/tasting notes, or fixing the timestamp (pulled_at). Changing grind_label re-derives the numeric grind position from the shot's grinder; changing yield/time/dose/tds keeps flow rate and extraction yield consistent automatically.
update_water
ChatGPTCorrect a registered water — its name, type, mineral figures or notes. Only the keys you send in fields change; send null to clear an optional field (an empty string also clears a text field, but not a number or a closed list). Required fields can be changed but not cleared. fields may be omitted when only reason is sent.
Capabilities
Writes
Example Prompts
Click any prompt to copy it.
Links
App Stats
44
Tools
3
Prompts
ChatGPT
Platforms
Category
Nutrition PlanningWorks with
ChatGPT
Data refreshed daily