reply-team/jason-ai / skills/business/sdr-operations
sdr-operations
Verifiedoperations/15-sending.yaml
View on GitHub# Family 15 — Sending capability and deliverability (34)
#
# A sending capability is a mailbox, a social account, a number or a business messaging identity —
# one noun, because every fact in this family holds for all of them. The family answers whether we
# can physically deliver, from what, how much, and what is stopping us right now.
#
# Schema, the derivation table and the meters: families.yaml.
family: 15
operations:
- name: sender.list
intent: >
"What can we send from?" — mailboxes, social accounts, numbers and business messaging
identities in one list.
reach: read
reversibility: reversible
approval: auto
approval_departs: false
before_repeating: nothing at stake
idempotency_key: none
per_item_results: not_applicable
cost: none
questions: []
- name: sender.get
intent: >
"Show me this sending capability in full" — identity, channel, configured pacing, message
defaults, pool membership, registration.
reach: read
reversibility: reversible
approval: auto
approval_departs: false
before_repeating: nothing at stake
idempotency_key: none
per_item_results: not_applicable
cost: none
questions: []
- name: sender.health
core: true
intent: >
"Can this sender send today, and if not why?" — connection, restrictions, warm-up position,
capacity budgets, and the channel's automation sanction.
reach: read
reversibility: reversible
approval: auto
approval_departs: false
before_repeating: nothing at stake
idempotency_key: none
per_item_results: required
cost: none
questions: []
- name: sender.connect
intent: >
"Attach a mailbox, social account, number or business messaging identity".
reach: control
reversibility: compensatable
approval: confirm_once
approval_departs: false
before_repeating: >
the sender's existence and state — a half-finished connection must read as `pending_setup`, not
as absent.
idempotency_key: required
accepts_collection: false
per_item_results: not_applicable
cost: none
questions: []
- name: sender.disconnect
intent: >
"Detach it" — refused while live enrolments or open conversations depend on it, unless a rebind
target is supplied.
reach: control
reversibility: compensatable
approval: confirm_once
approval_departs: false
before_repeating: >
the sender's state and what is still bound to it.
idempotency_key: required
accepts_collection: false
per_item_results: not_applicable
cost: none
questions: [15]
- name: sender.reauthorize
intent: >
"Restore a lapsed authorisation without losing the history behind it".
reach: control
reversibility: reversible
approval: auto
approval_departs: false
before_repeating: >
the sender's authorisation state and its last-checked time.
idempotency_key: required
accepts_collection: false
per_item_results: not_applicable
cost: none
questions: []
- name: sender.pause
intent: >
"Stop this capability sending, now" — the first move in every incident on every channel.
reach: control
reversibility: reversible
approval: auto
approval_departs: false
before_repeating: >
the sender's send state.
idempotency_key: required
per_item_results: required
cost: none
questions: []
- name: sender.resume
intent: >
"Let it send again" — refuses unless the warm-up position is valid or an override is recorded.
reach: control
reversibility: reversible
approval: confirm_once
approval_departs: true
approval_reason: >
raised from auto: sender.pause is protective, and the restart direction of a protective pair is
never auto. Platforms escalate on the frequency of restrictions, not their type, so resuming at
pre-restriction volume the moment a block lifts is the standard way to turn a temporary
restriction into a permanent one; a recorded restriction invalidates the current warm-up
position and this operation refuses while it is invalid unless a ramp restart is supplied or an
override is recorded with an actor and a reason.
before_repeating: >
the sender's send state, its warm-up position and whether that position was invalidated.
idempotency_key: required
per_item_results: required
cost: none
questions: []
- name: sender_limit.set
intent: >
"How much may this send, how fast, and with what spacing" — one decision, not three.
reach: control
reversibility: reversible
approval: confirm_once
approval_departs: true
approval_reason: >
raised from auto: one call declares how much this capability may send, how fast and with what
spacing, over limits somebody else may have declared — the read before repeating is the limits
in force and who declared each. The derivation table admits confirm_once where a call may
overwrite state someone else owns only in the compensatable column, so declaring it on a
reversible control is a raise.
before_repeating: >
the limits in force and who declared each.
idempotency_key: required
per_item_results: required
cost: none
questions: []
- name: warmup_plan.set
intent: >
"Define the ramp this sender follows, and whether it is running".
reach: control
reversibility: reversible
approval: confirm_once
approval_departs: true
approval_reason: >
raised from auto: warm-up is a state machine with a position, and that position is what
sender.resume refuses on — a recorded restriction invalidates it — so redefining the ramp
overwrites the state that governs whether the sender may send at all.
before_repeating: >
the current plan, position and whether the position is valid.
idempotency_key: required
accepts_collection: false
per_item_results: not_applicable
cost: none
questions: []
- name: message_element.set
intent: >
"The mandatory elements every message from this sender carries" — legal identification block,
postal address where required, the channel's required stop artefacts each separately (on email:
header, one-click endpoint, visible link), tracking domain.
reach: control
reversibility: reversible
approval: confirm_once
approval_departs: true
approval_reason: >
raised from auto: the call overwrites what every message sent afterwards is legally held to, and
the unsubscribe mechanism is three separate artefacts — a header, a machine-actionable endpoint
that actually honours a request, and a visible link in the body — each separately required,
configured and checked, where header presence proves only the first.
before_repeating: >
the elements currently in force.
idempotency_key: required
accepts_collection: false
per_item_results: not_applicable
cost: none
questions: []
- name: sender.reserve
intent: >
"Claim one unit of allowance so parallel work cannot spend it twice".
reach: control
reversibility: reversible
approval: auto
approval_departs: false
before_repeating: >
the reservation ledger for this sender and window.
idempotency_key: required
per_item_results: required
cost: none
questions: []
- name: sender.release
intent: >
"Give the claim back".
reach: control
reversibility: reversible
approval: auto
approval_departs: false
before_repeating: >
the reservation's state.
idempotency_key: required
per_item_results: required
cost: none
questions: []
- name: sender.summarize
intent: >
"This sender's numbers over a window" — failures by class, complaints with their publisher and
denominator, deferrals, throttles.
reach: read
reversibility: reversible
approval: auto
approval_departs: false
before_repeating: nothing at stake
idempotency_key: none
per_item_results: not_applicable
cost: none
questions: []
- name: restriction.record
intent: >
"The platform has restricted this sender" — including the ones we only learned about from a
refusal.
reach: control
reversibility: irreversible
approval: auto
approval_departs: true
protective_floor: true
approval_reason: >
the protective floor: recording an event that already happened outside our control derives auto
whatever its reversibility, and waiting for permission to record a restriction is the
destructive choice — the invalidated warm-up position and the recurrence count both hang off
this record.
before_repeating: >
the restriction ledger for this sender: an identical open restriction must not be double-counted
into the recurrence count.
idempotency_key: required
per_item_results: required
cost: none
questions: []
- name: restriction.appeal
intent: >
"Appeal it where the platform offers a route — and say plainly when it does not".
reach: act
reversibility: irreversible
approval: confirm_each
approval_departs: true
approval_reason: >
raised from confirm_once: an appeal is written content submitted under a named human's identity,
usually with one chance.
before_repeating: >
the appeal state for this restriction: none / submitted / decided.
idempotency_key: required
accepts_collection: false
per_item_results: not_applicable
cost: none
questions: []
- name: sender_pool.create
intent: >
"Group these capabilities so work can rotate across them".
reach: control
reversibility: compensatable
approval: confirm_once
approval_departs: false
before_repeating: >
the pool's existence and membership.
idempotency_key: required
accepts_collection: false
per_item_results: not_applicable
cost: none
questions: []
- name: sender_pool.update
intent: >
"Change who is in the pool and how work is routed across it" — including sticky per recipient,
which on conversational channels is correctness and not optimisation.
reach: control
reversibility: reversible
approval: confirm_once
approval_departs: true
approval_reason: >
raised from auto: the call carries the pool's membership and its routing policy, and rotation is
not a cure — where a budget's scope is the pool, the organisation or the recipient, rotating
across senders moves the problem, and sticky per recipient is correctness on conversational
channels rather than optimisation, so this overwrites state that live conversations depend on.
before_repeating: >
the pool's current membership and routing policy.
idempotency_key: required
per_item_results: required
cost: none
questions: []
- name: sender_pool.list
intent: >
"What pools exist, who is in them, and how do they route?"
reach: read
reversibility: reversible
approval: auto
approval_departs: false
before_repeating: nothing at stake
idempotency_key: none
per_item_results: not_applicable
cost: none
questions: []
- name: sender_pool.delete
intent: >
"Dissolve the pool" — refused while campaigns bind it unless a replacement target is supplied;
the capabilities themselves are untouched.
reach: control
reversibility: compensatable
approval: confirm_once
approval_departs: false
before_repeating: >
the pool's existence and what still binds it.
idempotency_key: required
accepts_collection: false
per_item_results: not_applicable
cost: none
questions: []
- name: sending_domain.register
intent: >
"Declare a domain we send from".
reach: control
reversibility: compensatable
approval: confirm_once
approval_departs: false
before_repeating: >
the domain's registration state.
idempotency_key: required
per_item_results: required
cost: none
questions: []
- name: sending_domain.list
intent: >
"Which domains do we send from, and what is each one's compliance state as of when?"
reach: read
reversibility: reversible
approval: auto
approval_departs: false
before_repeating: nothing at stake
idempotency_key: none
per_item_results: not_applicable
cost: none
questions: []
- name: sending_domain.retire
intent: >
"Stop treating this as a domain we send from" — refused while any sender on it has live
enrolments or open conversations, unless a rebind target is supplied.
reach: control
reversibility: compensatable
approval: confirm_once
approval_departs: false
before_repeating: >
the domain's state and what is still bound to it.
idempotency_key: required
accepts_collection: false
per_item_results: not_applicable
cost: none
questions: []
- name: authentication_requirement.list
intent: >
"What has to be published for this domain to authenticate" — publishing it is not ours to do.
reach: read
reversibility: reversible
approval: auto
approval_departs: false
before_repeating: nothing at stake
idempotency_key: none
per_item_results: not_applicable
cost: none
questions: []
- name: domain_compliance.check
intent: >
"Per publisher, per requirement: pass, fail or unknown — and name the failing requirement".
reach: read
reversibility: reversible
approval: auto
approval_departs: false
before_repeating: nothing at stake
idempotency_key: none
per_item_results: required
cost: none
questions: []
- name: domain_reputation.get
intent: >
"The named signals about this domain, each with its publisher, scale, window and denominator" —
including blocklist listings and the dataset each came from.
reach: read
reversibility: reversible
approval: auto
approval_departs: false
before_repeating: nothing at stake
idempotency_key:
value: required
conditional: true
detail: required where the read is metered.
per_item_results: required
cost:
value: metered
conditional: true
detail: none unless a signal comes from a paid dataset — then metered.
cost_basis: per dataset queried, where a signal comes from a paid dataset
meter: dataset
questions: []
- name: message_compliance.check
intent: >
"Check this rendered message against the requirements it will actually be held to" — required
elements, a working unsubscribe checked properly, and the reputation of every domain it links
to.
reach: read
reversibility: reversible
approval: auto
approval_departs: false
before_repeating: nothing at stake
idempotency_key: none
per_item_results: required
cost: none
questions: []
- name: placement_test.create
intent: >
"Send this to a monitored set of addresses and see where it lands".
reach: act
reversibility: irreversible
approval: confirm_each
approval_departs: false
before_repeating: >
the test ledger — a test already running for this message must not be started twice.
idempotency_key: required
accepts_collection: false
per_item_results: not_applicable
cost: metered
cost_basis: per test
meter: test
questions: []
- name: placement_test.get
intent: >
"What did the test say, per mailbox provider" — typed as simulated placement, never as our inbox
rate.
reach: read
reversibility: reversible
approval: auto
approval_departs: false
before_repeating: nothing at stake
idempotency_key: none
per_item_results: required
cost: none
questions: []
- name: messaging_registration.submit
intent: >
"Register the brand and the use case, because on this channel we may not send at all until it is
approved".
reach: act
reversibility: compensatable
approval: confirm_once
approval_departs: false
before_repeating: >
the registration's state: none / pending / approved / rejected.
idempotency_key: required
accepts_collection: false
per_item_results: not_applicable
cost:
value: metered
conditional: true
detail: metered where the channel charges for a registration; none where it does not.
cost_basis: per registration, where the channel charges for one
meter: registration
questions: []
- name: messaging_registration.get
intent: >
"What is the registration's status, and on refusal what must change?"
reach: read
reversibility: reversible
approval: auto
approval_departs: false
before_repeating: nothing at stake
idempotency_key: none
per_item_results: not_applicable
cost: none
questions: []
- name: approved_content.submit
intent: >
"Submit content for platform pre-approval, where the platform gates the content rather than the
sender".
reach: act
reversibility: compensatable
approval: confirm_once
approval_departs: false
before_repeating: >
the content's state: none / pending / approved / rejected / paused.
idempotency_key: required
accepts_collection: false
per_item_results: not_applicable
cost: none
questions: []
- name: approved_content.get
intent: >
"Approval status and quality state" — including whether the platform has paused it under us.
reach: read
reversibility: reversible
approval: auto
approval_departs: false
before_repeating: nothing at stake
idempotency_key: none
per_item_results: required
cost: none
questions: []
- name: approved_content.retire
intent: >
"Withdraw it from use" — the correct fix when it is the content that got flagged and not the
sender.
reach: control
reversibility: compensatable
approval: auto
approval_departs: false
before_repeating: >
the content's state and what is still bound to it.
idempotency_key: required
per_item_results: required
cost: none
questions: []
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |