Message Send
Execute one Slack channel post or source-bound reply on behalf of a principal.
The explicit plan runner can prepare provider-neutral authorization; delivery
requires the declared Slack operations. Email delivery belongs to the owning
mail capability, such as Nitrosend; social posting belongs to its account skill. The skill never receives a provider token, never lets
approved content drift before delivery, and calls a message delivered only
after the provider reads it back.
Runners
send (default): plans one channel post, requests approval only at the live
post, posts once, and requires independent readback.
plan: the provider-neutral authorization step. It binds the principal,
audience, content digest, consent basis, preflight checks and approval gate
into a runx/message-send.plan.v1 packet, and never delivers.
plan_post: validates a channel name, channel id, or exact
slack://workspace/channel destination, computes the message digest with
Runx's native data.digest, and binds that digest, the named principal and
the send intent through plan. It does not post and needs no approval.
plan_post_exact: for an operator that has already selected one exact
destination, principal and text under its own policy. It reuses the same
validation and digest and emits the same undelivered runx/message-send.notification_plan.v1 plan
without a second model judgment. Its send_plan.origin is
message-send.exact, not a claim that plan ran.
post: accepts only an exact notify_plan, the matching destination and
text, and a stable UUID idempotency key. It recomputes the digest, compares
the complete binding, stops unless exact native approval or a matching
standing notification authority is present, runs channel.post through the
configured binding, then reads the returned message locator back with
channel.post.read and seals only the bounded evidence fields.
reconcile_post: reads the exact selected message through channel.post.read,
with full content validation owned by the connector. It returns an independent
observation and explicitly leaves the original native effect status unchanged.
It never sends again or treats a timestamp/preview match as settled custody.
reply: delivers one admitted reply to a bound conversation. The operator
supplies the exact source packet, thread, draft, provider and native turn
identity. The skill holds unless the source proves complete coverage, then
asks for exact approval, runs thread.reply on the chosen connector, and
requires thread.reply.read to resolve a stable message locator with the
matching content digest. An accepted or queued request alone leaves the work
held for reconciliation.
A channel post and a source-bound reply keep their own input checks: a post is
bound to a destination and a plan, a reply to a source packet and a turn.
The binding may be local, self-hosted, third-party or Runx-hosted. The skill
never receives a token, constructs a raw HTTP request, owns a tenant, or stores
connector credentials in its inputs or receipts.
When to use this skill
- An agent needs to post one bounded message, or reply in one conversation, on
behalf of a user, team, brand, account or service.
- An operator explicitly wants an authorization plan and selects
plan.
- The workflow must prove the intended audience, content, consent basis and
approval decision before delivery.
When not to use this skill
- To write copy only. Use a drafting or brand-voice skill unless delivery is in
scope.
- To claim delivery from a plan; only provider readback is terminal evidence.
- To send without a named principal and audience.
- To bypass unsubscribe, consent, suppression, preflight, legal or human
approval gates.
Procedure for plan
- Identify the principal being represented and the provider account or surface.
- Classify the send:
transactional, campaign, flow_step, support_reply,
outreach, status, or internal.
- Bind channel and audience. Audience must be a named recipient, list, segment,
support thread, channel, or scoped all-contacts decision; never an implicit
broad default.
- Bind content by digest or stable draft reference. Do not approve mutable
content by prose summary alone.
- Check consent, unsubscribe, suppression, compliance, preflight, and provider
readiness. Missing evidence becomes a blocker.
- Decide the gate:
- drafts, previews, and test sends may proceed without live-delivery
approval when provider policy permits them;
- customer, public, audience, or live sends require explicit approval;
- billing/account mutation is outside this skill and needs its own gate.
- Produce the smallest provider-neutral
send_plan that execution can consume
without widening authority.
- Return
needs_input for missing principal, audience, content digest,
consent basis, or provider readiness; return refused for gate bypass.
Edge cases and stop conditions
- No principal: return
needs_input; the agent cannot speak as an unnamed
actor.
- No audience: return
needs_input; do not default to all contacts or a
whole channel.
- All contacts or broad audience: require explicit reconfirmation and a
stricter preflight block.
- Mutable content: return
needs_input until content is digest-bound.
- Missing consent or unsubscribe path: block live delivery.
- Plan, destination or content-digest drift: return an explicit held outcome
with no completion evidence; prepare and approve the changed content.
- Approval denied or absent: do not deliver.
- Missing, ambiguous, wrong-provider or insufficient-scope binding: refuse
rather than fall back to a raw token.
- Ambiguous mutation outcome: resume with the same idempotency key, or
reconcile; never create a new key to escape uncertainty. Changed content
never inherits the old approval.
- Raw credentials or contact dumps: redact; if redaction would remove the
evidence needed to decide, return
needs_input.
Outputs
send_plan is the runx/message-send.plan.v1 packet: the decision
(ready, needs_input, denied or refused), the principal, provider,
send class, channel, audience, digest-bound content, the gates the send
needs, blockers, provider actions, evidence locators and the success
checkpoint.
notify_plan is the runx/message-send.notification_plan.v1 packet, the
exact post post delivers and reconcile_post can independently observe.
- A reply ends in a
runx.work_outcome.v1 packet for its host.
A post is delivered only when the sealed runx.provider.operation.v1 packet
identifies the operation and carries the provider-returned message locator,
conversation locator, content digest and occurrence time, and the follow-up
readback resolves that exact locator.
Exact text includes leading spaces and final line breaks. The native input and
provider boundary accept up to Slack's 40,000-character text limit; keep ordinary
messages short. Truncated or split content is not successful exact readback. After an ambiguous send, retain the original pending run and native effect.
reconcile_post can observe a selected message, but its result is not authority
to resend or claim that the original native effect has settled. Known provider
acknowledgements resume through the original native readback path.
Worked example
An operator plans "release 2.4 is verified" for one internal channel. Approval
binds that exact text and destination. If the text changes to include another
claim, post refuses the digest mismatch. With a matching plan and one valid
Slack binding, the provider posts and returns the message id; the follow-up
readback resolves that id, and those provider observations, not the local plan,
prove delivery.
Inputs
send: channel_locator, text, principal, idempotency_key, and
optional provider_context.
plan: objective and principal, with optional provider_context,
audience, content, consent_basis, operator_context and
connector.
post: the notify_plan, the identical locator and text, and a stable
idempotency key.
reply: source_context, turn_id, expected_provider, thread_locator
and draft_message.
1---2name: message-send3description: Send one digest-bound message on behalf of a principal, posting to a channel or replying in a bound conversation through a compatible connector, after exact approval and independent provider readback.4metadata:5 source: https://skillmd.com/skills/runxhq/message-send6---78# Message Send910Execute one Slack channel post or source-bound reply on behalf of a principal.11The explicit `plan` runner can prepare provider-neutral authorization; delivery12requires the declared Slack operations. Email delivery belongs to the owning13mail capability, such as Nitrosend; social posting belongs to its account skill. The skill never receives a provider token, never lets14approved content drift before delivery, and calls a message delivered only15after the provider reads it back.1617## Runners1819- `send` (default): plans one channel post, requests approval only at the live20 post, posts once, and requires independent readback.21- `plan`: the provider-neutral authorization step. It binds the principal,22 audience, content digest, consent basis, preflight checks and approval gate23 into a `runx/message-send.plan.v1` packet, and never delivers.24- `plan_post`: validates a channel name, channel id, or exact25 `slack://workspace/channel` destination, computes the message digest with26 Runx's native `data.digest`, and binds that digest, the named principal and27 the send intent through `plan`. It does not post and needs no approval.28- `plan_post_exact`: for an operator that has already selected one exact29 destination, principal and text under its own policy. It reuses the same30 validation and digest and emits the same undelivered `runx/message-send.notification_plan.v1` plan31 without a second model judgment. Its `send_plan.origin` is32 `message-send.exact`, not a claim that `plan` ran.33- `post`: accepts only an exact `notify_plan`, the matching destination and34 text, and a stable UUID idempotency key. It recomputes the digest, compares35 the complete binding, stops unless exact native approval or a matching36 standing notification authority is present, runs `channel.post` through the37 configured binding, then reads the returned message locator back with38 `channel.post.read` and seals only the bounded evidence fields.39- `reconcile_post`: reads the exact selected message through `channel.post.read`,40 with full content validation owned by the connector. It returns an independent41 observation and explicitly leaves the original native effect status unchanged.42 It never sends again or treats a timestamp/preview match as settled custody.43- `reply`: delivers one admitted reply to a bound conversation. The operator44 supplies the exact source packet, thread, draft, provider and native turn45 identity. The skill holds unless the source proves complete coverage, then46 asks for exact approval, runs `thread.reply` on the chosen connector, and47 requires `thread.reply.read` to resolve a stable message locator with the48 matching content digest. An accepted or queued request alone leaves the work49 held for reconciliation.5051A channel post and a source-bound reply keep their own input checks: a post is52bound to a destination and a plan, a reply to a source packet and a turn.5354The binding may be local, self-hosted, third-party or Runx-hosted. The skill55never receives a token, constructs a raw HTTP request, owns a tenant, or stores56connector credentials in its inputs or receipts.5758## When to use this skill5960- An agent needs to post one bounded message, or reply in one conversation, on61 behalf of a user, team, brand, account or service.62- An operator explicitly wants an authorization plan and selects `plan`.63- The workflow must prove the intended audience, content, consent basis and64 approval decision before delivery.6566## When not to use this skill6768- To write copy only. Use a drafting or brand-voice skill unless delivery is in69 scope.70- To claim delivery from a plan; only provider readback is terminal evidence.71- To send without a named principal and audience.72- To bypass unsubscribe, consent, suppression, preflight, legal or human73 approval gates.7475## Procedure for `plan`76771. Identify the principal being represented and the provider account or surface.782. Classify the send: `transactional`, `campaign`, `flow_step`, `support_reply`,79 `outreach`, `status`, or `internal`.803. Bind channel and audience. Audience must be a named recipient, list, segment,81 support thread, channel, or scoped all-contacts decision; never an implicit82 broad default.834. Bind content by digest or stable draft reference. Do not approve mutable84 content by prose summary alone.855. Check consent, unsubscribe, suppression, compliance, preflight, and provider86 readiness. Missing evidence becomes a blocker.876. Decide the gate:88 - drafts, previews, and test sends may proceed without live-delivery89 approval when provider policy permits them;90 - customer, public, audience, or live sends require explicit approval;91 - billing/account mutation is outside this skill and needs its own gate.927. Produce the smallest provider-neutral `send_plan` that execution can consume93 without widening authority.948. Return `needs_input` for missing principal, audience, content digest,95 consent basis, or provider readiness; return `refused` for gate bypass.9697## Edge cases and stop conditions9899- **No principal:** return `needs_input`; the agent cannot speak as an unnamed100 actor.101- **No audience:** return `needs_input`; do not default to all contacts or a102 whole channel.103- **All contacts or broad audience:** require explicit reconfirmation and a104 stricter preflight block.105- **Mutable content:** return `needs_input` until content is digest-bound.106- **Missing consent or unsubscribe path:** block live delivery.107- **Plan, destination or content-digest drift:** return an explicit held outcome108 with no completion evidence; prepare and approve the changed content.109- **Approval denied or absent:** do not deliver.110- **Missing, ambiguous, wrong-provider or insufficient-scope binding:** refuse111 rather than fall back to a raw token.112- **Ambiguous mutation outcome:** resume with the same idempotency key, or113 reconcile; never create a new key to escape uncertainty. Changed content114 never inherits the old approval.115- **Raw credentials or contact dumps:** redact; if redaction would remove the116 evidence needed to decide, return `needs_input`.117118## Outputs119120- `send_plan` is the `runx/message-send.plan.v1` packet: the decision121 (`ready`, `needs_input`, `denied` or `refused`), the principal, provider,122 send class, channel, audience, digest-bound content, the gates the send123 needs, blockers, provider actions, evidence locators and the success124 checkpoint.125- `notify_plan` is the `runx/message-send.notification_plan.v1` packet, the126 exact post `post` delivers and `reconcile_post` can independently observe.127- A reply ends in a `runx.work_outcome.v1` packet for its host.128129A post is delivered only when the sealed `runx.provider.operation.v1` packet130identifies the operation and carries the provider-returned message locator,131conversation locator, content digest and occurrence time, and the follow-up132readback resolves that exact locator.133134Exact text includes leading spaces and final line breaks. The native input and135provider boundary accept up to Slack's 40,000-character text limit; keep ordinary136messages short. Truncated or split content is not successful exact readback. After an ambiguous send, retain the original pending run and native effect.137`reconcile_post` can observe a selected message, but its result is not authority138to resend or claim that the original native effect has settled. Known provider139acknowledgements resume through the original native readback path.140141## Worked example142143An operator plans "release 2.4 is verified" for one internal channel. Approval144binds that exact text and destination. If the text changes to include another145claim, `post` refuses the digest mismatch. With a matching plan and one valid146Slack binding, the provider posts and returns the message id; the follow-up147readback resolves that id, and those provider observations, not the local plan,148prove delivery.149150## Inputs151152- `send`: `channel_locator`, `text`, `principal`, `idempotency_key`, and153 optional `provider_context`.154- `plan`: `objective` and `principal`, with optional `provider_context`,155 `audience`, `content`, `consent_basis`, `operator_context` and156 `connector`.157- `post`: the `notify_plan`, the identical locator and text, and a stable158 idempotency key.159- `reply`: `source_context`, `turn_id`, `expected_provider`, `thread_locator`160 and `draft_message`.