Outstand

WORKING SLICE
Expose a small authenticated HTTP publishing API with text/media validation and independent adapters for approved networks. Provide create/status/cancel endpoints and normalize available metrics with null for unavailable fields.

Build this scoped Outstand-inspired workflow with a documented data model and visible failure states.

Architecture
- Node.js, TypeScript, Fastify and SQLite for a private publishing API and scheduler.
- Schema-validated normalized posts with separate official-network adapters; no interactive web frontend is required.

Prerequisites and limits
Node, a private authenticated API endpoint, a persistent SQLite directory and access credentials for each explicitly enabled network. Validate current API permissions and quotas.
Outside this release: pre-approved platform apps for Instagram, TikTok, YouTube and LinkedIn; eleven maintained integrations, kept alive as platforms change

Data model and correctness
API keys, normalized posts, target accounts, capability declarations, job IDs and per-target receipts
Invariant: An idempotency key is scoped to the caller and immutable request body; partial success never becomes a whole-job success placeholder.
Track each target independently with pending, publishing, delivered, failed and unknown states. Reconcile a timeout using provider receipts before retrying; respect rate-limit backoff. An optional stdio MCP adapter may expose the same authenticated create/status operations with identical approval requirements.

Security and privacy
Keep OAuth tokens server-side and encrypt persistent tokens with a separately managed key. Publication requires an explicit approved revision and target account.

Recovery and export
Use a consistent SQLite backup and an attachment manifest. Export portable JSON/CSV, then restore to a new directory without overwriting the original data.

Implementation order
1. Phase 1 — Scope and fixtures. Implement this bounded workflow: Expose a small authenticated HTTP publishing API with text/media validation and independent adapters for approved networks. Provide create/status/cancel endpoints and normalize available metrics with null for unavailable fields. Record prerequisites, select representative user-owned fixtures and document the unsupported features: pre-approved platform apps for Instagram, TikTok, YouTube and LinkedIn; eleven maintained integrations, kept alive as platforms change
2. Phase 2 — Durable model. Model API keys, normalized posts, target accounts, capability declarations, job IDs and per-target receipts Add migrations or a versioned document format, explicit validation, stable IDs and a visible import-error report. Preserve this rule: An idempotency key is scoped to the caller and immutable request body; partial success never becomes a whole-job success placeholder.
3. Phase 3 — Complete the first useful path. Implement the workflow's input, review and output interface, with clear controls and explicit empty/error states. Track each target independently with pending, publishing, delivered, failed and unknown states. Reconcile a timeout using provider receipts before retrying; respect rate-limit backoff. An optional stdio MCP adapter may expose the same authenticated create/status operations with identical approval requirements.
4. Phase 4 — Permissions and integration failure. Keep OAuth tokens server-side and encrypt persistent tokens with a separately managed key. Publication requires an explicit approved revision and target account. Request integration credentials and permissions only for the enabled feature; show a disconnected state instead of mock results.
5. Phase 5 — Portable handoff. Use a consistent SQLite backup and an attachment manifest. Export portable JSON/CSV, then restore to a new directory without overwriting the original data. Include setup, operating limits, fixture walkthrough and shutdown/restart instructions in the README.
6. Phase 6 — Acceptance scenarios. Replay a request and receive the same job; one failed target can retry without reposting to delivered targets, and missing impressions remain null. Repeat the workflow after restart and with a denied permission or unavailable dependency; show recoverable failure rather than a success placeholder.

Acceptance
Replay a request and receive the same job; one failed target can retry without reposting to delivered targets, and missing impressions remain null.
Use real source data or clearly labeled fixtures. Explain unsupported input and provider failures; do not fabricate analytics, delivery receipts, accuracy claims or security guarantees.

Optional agent guidance
Optional external skill: [sharp-edges](https://github.com/trailofbits/skills/blob/main/plugins/sharp-edges/skills/sharp-edges/SKILL.md) — Review security-sensitive APIs and configuration for dangerous defaults and easy-to-misuse interfaces. Review its instructions and compatibility before use; it does not grant deployment, data-access or publication permission.
Optional external skill: [mcp-builder](https://github.com/anthropics/skills/blob/main/skills/mcp-builder/SKILL.md) — Build MCP servers that expose external APIs through tools using Python FastMCP or the Node/TypeScript MCP SDK. Review its instructions and compatibility before use; it does not grant deployment, data-access or publication permission.
Project rule — data model: API keys, normalized posts, target accounts, capability declarations, job IDs and per-target receipts
Project rule — preserve this invariant: An idempotency key is scoped to the caller and immutable request body; partial success never becomes a whole-job success placeholder.
Project rule — acceptance evidence: Replay a request and receive the same job; one failed target can retry without reposting to delivered targets, and missing impressions remain null.

# AGENTS.md — Build guide for Outstand

## Project scope
Expose a small authenticated HTTP publishing API with text/media validation and independent adapters for approved networks. Provide create/status/cancel endpoints and normalize available metrics with null for unavailable fields.

Catalogue verdict: kinda. Writing your own posting service is a weekend build, and for the networks with open APIs (Bluesky, Mastodon, Telegram) it genuinely is: a queue, a cron tick, one adapter per network, done. The wall is not code. Instagram, TikTok, YouTube, LinkedIn, and Pinterest each require your own developer app and a platform review before they will accept a single post, X meters writes on a paid tier, and then eleven integrations keep changing under you: token refresh, media specs, rate-limit budgets, deprecated endpoints. You can build the API. You cannot one-shot permission to use it.
Use the implementation prompt below to define the deliverable. Complete each phase's acceptance checks before extending the scope.

## Working agreement
- Inspect the repository and its existing instructions before choosing paths, dependencies or commands. Keep one coherent stack and explain changes to the proposed architecture.
- Plan a vertical slice that accepts a real input and produces the useful output described below. Persist only the state the prompt calls for; respect memory-only and upstream-managed workflows. Use fixtures only when they are clearly labelled.
- After scaffolding, document the actual install, development, check and build commands in README and keep them synchronized with the package or project manifest. Do not report commands as successful unless they ran.
- Work in small steps. At handoff, list implemented flows, checks actually performed, remaining blockers, and any credentials or provider setup the owner must supply.
- Do not publish, spend money, contact customers, delete source data or run irreversible migrations without the project owner's authorization.

## Prerequisites
- Node, a private authenticated API endpoint, a persistent SQLite directory and access credentials for each explicitly enabled network. Validate current API permissions and quotas.
- Implementation components: Node.js, TypeScript, Fastify and SQLite for a private publishing API and scheduler. Schema-validated normalized posts with separate official-network adapters; no interactive web frontend is required.
- Scope boundary: pre-approved platform apps for Instagram, TikTok, YouTube and LinkedIn; eleven maintained integrations, kept alive as platforms change

## Stack and architecture
- Node.js, TypeScript, Fastify and SQLite for a private publishing API and scheduler.
- Schema-validated normalized posts with separate official-network adapters; no interactive web frontend is required.
- Domain model: API keys, normalized posts, target accounts, capability declarations, job IDs and per-target receipts

## Security and data integrity
- Keep OAuth tokens server-side and encrypt persistent tokens with a separately managed key. Publication requires an explicit approved revision and target account.
- Correctness boundary: An idempotency key is scoped to the caller and immutable request body; partial success never becomes a whole-job success placeholder.
- Track each target independently with pending, publishing, delivered, failed and unknown states. Reconcile a timeout using provider receipts before retrying; respect rate-limit backoff. An optional stdio MCP adapter may expose the same authenticated create/status operations with identical approval requirements.
- Use a consistent SQLite backup and an attachment manifest. Export portable JSON/CSV, then restore to a new directory without overwriting the original data.

## Agent implementation rules
- Project rule — data model: API keys, normalized posts, target accounts, capability declarations, job IDs and per-target receipts
- Project rule — preserve this invariant: An idempotency key is scoped to the caller and immutable request body; partial success never becomes a whole-job success placeholder.
- Project rule — acceptance evidence: Replay a request and receive the same job; one failed target can retry without reposting to delivered targets, and missing impressions remain null.

## Optional agent skills and references
- Optional external skill: [sharp-edges](https://github.com/trailofbits/skills/blob/main/plugins/sharp-edges/skills/sharp-edges/SKILL.md) — Review security-sensitive APIs and configuration for dangerous defaults and easy-to-misuse interfaces. Review its instructions and compatibility before use; it does not grant deployment, data-access or publication permission.
- Optional external skill: [mcp-builder](https://github.com/anthropics/skills/blob/main/skills/mcp-builder/SKILL.md) — Build MCP servers that expose external APIs through tools using Python FastMCP or the Node/TypeScript MCP SDK. Review its instructions and compatibility before use; it does not grant deployment, data-access or publication permission.

Read the linked SKILL.md and its dependencies before adding a skill. Select only the skills matching this project's runtime and task; their documentation does not supply API access, credentials or approval to perform external actions. Pin the reviewed revision where the tool supports it. Follow the chosen agent's documented project-level installation mechanism.

## Distribution ideas
These are optional planning notes. Obtain the owner's approval before publishing or contacting anyone.
- Demonstrate the actual Outstand-inspired workflow with owned or clearly labeled sample data: Expose a small authenticated HTTP publishing API with text/media validation and independent adapters for approved networks. Provide create/status/cancel endpoints and normalize available metrics with null for unavailable fields.
- Publish a reproducible walkthrough with this observable result: Replay a request and receive the same job; one failed target can retry without reposting to delivered targets, and missing impressions remain null.
- Explain who can operate this scoped tool, its setup and ongoing costs, and these remaining product gaps: pre-approved platform apps for Instagram, TikTok, YouTube and LinkedIn; eleven maintained integrations, kept alive as platforms change Avoid guaranteed savings, performance scores or implied endorsement.

## Engineering roadmap
1. Phase 1 — Scope and fixtures. Implement this bounded workflow: Expose a small authenticated HTTP publishing API with text/media validation and independent adapters for approved networks. Provide create/status/cancel endpoints and normalize available metrics with null for unavailable fields. Record prerequisites, select representative user-owned fixtures and document the unsupported features: pre-approved platform apps for Instagram, TikTok, YouTube and LinkedIn; eleven maintained integrations, kept alive as platforms change
2. Phase 2 — Durable model. Model API keys, normalized posts, target accounts, capability declarations, job IDs and per-target receipts Add migrations or a versioned document format, explicit validation, stable IDs and a visible import-error report. Preserve this rule: An idempotency key is scoped to the caller and immutable request body; partial success never becomes a whole-job success placeholder.
3. Phase 3 — Complete the first useful path. Implement the workflow's input, review and output interface, with clear controls and explicit empty/error states. Track each target independently with pending, publishing, delivered, failed and unknown states. Reconcile a timeout using provider receipts before retrying; respect rate-limit backoff. An optional stdio MCP adapter may expose the same authenticated create/status operations with identical approval requirements.
4. Phase 4 — Permissions and integration failure. Keep OAuth tokens server-side and encrypt persistent tokens with a separately managed key. Publication requires an explicit approved revision and target account. Request integration credentials and permissions only for the enabled feature; show a disconnected state instead of mock results.
5. Phase 5 — Portable handoff. Use a consistent SQLite backup and an attachment manifest. Export portable JSON/CSV, then restore to a new directory without overwriting the original data. Include setup, operating limits, fixture walkthrough and shutdown/restart instructions in the README.
6. Phase 6 — Acceptance scenarios. Replay a request and receive the same job; one failed target can retry without reposting to delivered targets, and missing impressions remain null. Repeat the workflow after restart and with a denied permission or unavailable dependency; show recoverable failure rather than a success placeholder.

## Paid-product capabilities outside this build
- pre-approved platform apps for Instagram, TikTok, YouTube and LinkedIn
- eleven maintained integrations, kept alive as platforms change
- automatic token refresh and rate-limit queueing
- per-network media transcoding from a single upload
- cross-platform analytics in one shape
- the MCP server that hands all of it to an AI agent

## Implementation prompt
WORKING SLICE
Expose a small authenticated HTTP publishing API with text/media validation and independent adapters for approved networks. Provide create/status/cancel endpoints and normalize available metrics with null for unavailable fields.

Build this scoped Outstand-inspired workflow with a documented data model and visible failure states.

Architecture
- Node.js, TypeScript, Fastify and SQLite for a private publishing API and scheduler.
- Schema-validated normalized posts with separate official-network adapters; no interactive web frontend is required.

Prerequisites and limits
Node, a private authenticated API endpoint, a persistent SQLite directory and access credentials for each explicitly enabled network. Validate current API permissions and quotas.
Outside this release: pre-approved platform apps for Instagram, TikTok, YouTube and LinkedIn; eleven maintained integrations, kept alive as platforms change

Data model and correctness
API keys, normalized posts, target accounts, capability declarations, job IDs and per-target receipts
Invariant: An idempotency key is scoped to the caller and immutable request body; partial success never becomes a whole-job success placeholder.
Track each target independently with pending, publishing, delivered, failed and unknown states. Reconcile a timeout using provider receipts before retrying; respect rate-limit backoff. An optional stdio MCP adapter may expose the same authenticated create/status operations with identical approval requirements.

Security and privacy
Keep OAuth tokens server-side and encrypt persistent tokens with a separately managed key. Publication requires an explicit approved revision and target account.

Recovery and export
Use a consistent SQLite backup and an attachment manifest. Export portable JSON/CSV, then restore to a new directory without overwriting the original data.

Implementation order
1. Phase 1 — Scope and fixtures. Implement this bounded workflow: Expose a small authenticated HTTP publishing API with text/media validation and independent adapters for approved networks. Provide create/status/cancel endpoints and normalize available metrics with null for unavailable fields. Record prerequisites, select representative user-owned fixtures and document the unsupported features: pre-approved platform apps for Instagram, TikTok, YouTube and LinkedIn; eleven maintained integrations, kept alive as platforms change
2. Phase 2 — Durable model. Model API keys, normalized posts, target accounts, capability declarations, job IDs and per-target receipts Add migrations or a versioned document format, explicit validation, stable IDs and a visible import-error report. Preserve this rule: An idempotency key is scoped to the caller and immutable request body; partial success never becomes a whole-job success placeholder.
3. Phase 3 — Complete the first useful path. Implement the workflow's input, review and output interface, with clear controls and explicit empty/error states. Track each target independently with pending, publishing, delivered, failed and unknown states. Reconcile a timeout using provider receipts before retrying; respect rate-limit backoff. An optional stdio MCP adapter may expose the same authenticated create/status operations with identical approval requirements.
4. Phase 4 — Permissions and integration failure. Keep OAuth tokens server-side and encrypt persistent tokens with a separately managed key. Publication requires an explicit approved revision and target account. Request integration credentials and permissions only for the enabled feature; show a disconnected state instead of mock results.
5. Phase 5 — Portable handoff. Use a consistent SQLite backup and an attachment manifest. Export portable JSON/CSV, then restore to a new directory without overwriting the original data. Include setup, operating limits, fixture walkthrough and shutdown/restart instructions in the README.
6. Phase 6 — Acceptance scenarios. Replay a request and receive the same job; one failed target can retry without reposting to delivered targets, and missing impressions remain null. Repeat the workflow after restart and with a denied permission or unavailable dependency; show recoverable failure rather than a success placeholder.

Acceptance
Replay a request and receive the same job; one failed target can retry without reposting to delivered targets, and missing impressions remain null.
Use real source data or clearly labeled fixtures. Explain unsupported input and provider failures; do not fabricate analytics, delivery receipts, accuracy claims or security guarantees.

Optional agent guidance
Optional external skill: [sharp-edges](https://github.com/trailofbits/skills/blob/main/plugins/sharp-edges/skills/sharp-edges/SKILL.md) — Review security-sensitive APIs and configuration for dangerous defaults and easy-to-misuse interfaces. Review its instructions and compatibility before use; it does not grant deployment, data-access or publication permission.
Optional external skill: [mcp-builder](https://github.com/anthropics/skills/blob/main/skills/mcp-builder/SKILL.md) — Build MCP servers that expose external APIs through tools using Python FastMCP or the Node/TypeScript MCP SDK. Review its instructions and compatibility before use; it does not grant deployment, data-access or publication permission.
Project rule — data model: API keys, normalized posts, target accounts, capability declarations, job IDs and per-target receipts
Project rule — preserve this invariant: An idempotency key is scoped to the caller and immutable request body; partial success never becomes a whole-job success placeholder.
Project rule — acceptance evidence: Replay a request and receive the same job; one failed target can retry without reposting to delivered targets, and missing impressions remain null.

## Completion evidence
Demonstrate the prompt's acceptance scenarios against the scoped workflow. Include setup from a clean checkout and failure recovery. Check persistence across restart and export/restore only for the state the prompt says to store; for memory-only tools, confirm that temporary content is discarded as specified. Record actual results and remaining limitations. A detailed plan alone does not establish a working replacement.