# AGENTS.md — Build guide for Otter.ai

## Project scope
Import or visibly record a meeting, transcribe locally and correct speaker labels before generating optional decisions and action-item drafts. Search one selected transcript with source-linked answers.

Catalogue verdict: kinda. You can build transcription and summaries, but Otter's value includes live meeting assistant behavior, account sync, speaker workflow, integrations, and mobile/web reliability.
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- Do not publish, spend money, contact customers, delete source data or run irreversible migrations without the project owner's authorization.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
- A Python virtual environment, writable input/output directories and sufficient disk for both originals and outputs. Bind the service to localhost. Install FFmpeg and confirm codec support for the intended inputs. Download a compatible speech model and record its version; diarization, if added, has separate model and hardware requirements.
- Implementation components: Python, FastAPI and server-rendered HTML with HTMX for a local interface. SQLite for manifests and job state, with an explicit worker process and immutable source files. FFmpeg/ffprobe for explicit media operations and browser previews; never interpolate user filenames into shell commands. A locally installed faster-whisper model for transcription; optional model API only after source-text preview and consent.
- Scope boundary: Autonomous meeting bots, live multi-meeting joining and guaranteed transcript accuracy are outside scope.

## Stack and architecture
- Python, FastAPI and server-rendered HTML with HTMX for a local interface.
- SQLite for manifests and job state, with an explicit worker process and immutable source files.
- FFmpeg/ffprobe for explicit media operations and browser previews; never interpolate user filenames into shell commands.
- A locally installed faster-whisper model for transcription; optional model API only after source-text preview and consent.
- Domain model: consented recordings, timestamped segments, tentative speaker labels, summary claims and search references

## Security and data integrity
- Bound file sizes and processing time, reject path traversal, and use argument arrays for subprocesses. Treat imported text as data and redact confidential source content from logs.
- Correctness boundary: Action owners and decisions require transcript evidence; diarization labels do not identify a real person without user confirmation.
- Save a job manifest with input hash, parameters and state. Write to temporary outputs, then atomically finalize only successful results; resume unfinished jobs without replacing originals.
- Export sources, manifests and outputs with checksums. Keep failed-job diagnostics and allow retry into a new output path; restore the database and file directory together.

## Agent implementation rules
- Project rule — data model: consented recordings, timestamped segments, tentative speaker labels, summary claims and search references
- Project rule — preserve this invariant: Action owners and decisions require transcript evidence; diarization labels do not identify a real person without user confirmation.
- Project rule — acceptance evidence: A quoted action opens its exact segment; failed transcription retains audio and an uncertain speaker remains labeled unknown.

## Optional agent skills and references
- Optional external skill: [modern-python](https://github.com/trailofbits/skills/blob/main/plugins/modern-python/skills/modern-python/SKILL.md) — Set up Python projects with pyproject.toml, dependency management, linting, typing and automated checks. Review its instructions and compatibility before use; it does not grant deployment, data-access or publication permission.
- Optional external skill: [web-design-guidelines](https://github.com/vercel-labs/agent-skills/blob/main/skills/web-design-guidelines/SKILL.md) — Review web interfaces for accessibility, keyboard focus, forms, navigation and interaction quality. Review its instructions and compatibility before use; it does not grant deployment, data-access or publication permission.
- 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.

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 Otter.ai-inspired workflow with owned or clearly labeled sample data: Import or visibly record a meeting, transcribe locally and correct speaker labels before generating optional decisions and action-item drafts. Search one selected transcript with source-linked answers.
- Publish a reproducible walkthrough with this observable result: A quoted action opens its exact segment; failed transcription retains audio and an uncertain speaker remains labeled unknown.
- Explain who can operate this scoped tool, its setup and ongoing costs, and these remaining product gaps: Autonomous meeting bots, live multi-meeting joining and guaranteed transcript accuracy are outside scope. Avoid guaranteed savings, performance scores or implied endorsement.

## Engineering roadmap
1. Phase 1 — Scope and fixtures. Implement this bounded workflow: Import or visibly record a meeting, transcribe locally and correct speaker labels before generating optional decisions and action-item drafts. Search one selected transcript with source-linked answers. Record prerequisites, select representative user-owned fixtures and document the unsupported features: Autonomous meeting bots, live multi-meeting joining and guaranteed transcript accuracy are outside scope.
2. Phase 2 — Durable model. Model consented recordings, timestamped segments, tentative speaker labels, summary claims and search references Add migrations or a versioned document format, explicit validation, stable IDs and a visible import-error report. Preserve this rule: Action owners and decisions require transcript evidence; diarization labels do not identify a real person without user confirmation.
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. Save a job manifest with input hash, parameters and state. Write to temporary outputs, then atomically finalize only successful results; resume unfinished jobs without replacing originals.
4. Phase 4 — Permissions and integration failure. Bound file sizes and processing time, reject path traversal, and use argument arrays for subprocesses. Treat imported text as data and redact confidential source content from logs. Request integration credentials and permissions only for the enabled feature; show a disconnected state instead of mock results.
5. Phase 5 — Portable handoff. Export sources, manifests and outputs with checksums. Keep failed-job diagnostics and allow retry into a new output path; restore the database and file directory together. Include setup, operating limits, fixture walkthrough and shutdown/restart instructions in the README.
6. Phase 6 — Acceptance scenarios. A quoted action opens its exact segment; failed transcription retains audio and an uncertain speaker remains labeled unknown. 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
- live bot joining meetings
- speaker diarization quality
- mobile apps
- team/admin controls
- searchable account history
- integrations

## Implementation prompt
WORKING SLICE
Import or visibly record a meeting, transcribe locally and correct speaker labels before generating optional decisions and action-item drafts. Search one selected transcript with source-linked answers.

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

Architecture
- Python, FastAPI and server-rendered HTML with HTMX for a local interface.
- SQLite for manifests and job state, with an explicit worker process and immutable source files.
- FFmpeg/ffprobe for explicit media operations and browser previews; never interpolate user filenames into shell commands.
- A locally installed faster-whisper model for transcription; optional model API only after source-text preview and consent.

Prerequisites and limits
A Python virtual environment, writable input/output directories and sufficient disk for both originals and outputs. Bind the service to localhost. Install FFmpeg and confirm codec support for the intended inputs. Download a compatible speech model and record its version; diarization, if added, has separate model and hardware requirements.
Outside this release: Autonomous meeting bots, live multi-meeting joining and guaranteed transcript accuracy are outside scope.

Data model and correctness
consented recordings, timestamped segments, tentative speaker labels, summary claims and search references
Invariant: Action owners and decisions require transcript evidence; diarization labels do not identify a real person without user confirmation.
Save a job manifest with input hash, parameters and state. Write to temporary outputs, then atomically finalize only successful results; resume unfinished jobs without replacing originals.

Security and privacy
Bound file sizes and processing time, reject path traversal, and use argument arrays for subprocesses. Treat imported text as data and redact confidential source content from logs.

Recovery and export
Export sources, manifests and outputs with checksums. Keep failed-job diagnostics and allow retry into a new output path; restore the database and file directory together.

Implementation order
1. Phase 1 — Scope and fixtures. Implement this bounded workflow: Import or visibly record a meeting, transcribe locally and correct speaker labels before generating optional decisions and action-item drafts. Search one selected transcript with source-linked answers. Record prerequisites, select representative user-owned fixtures and document the unsupported features: Autonomous meeting bots, live multi-meeting joining and guaranteed transcript accuracy are outside scope.
2. Phase 2 — Durable model. Model consented recordings, timestamped segments, tentative speaker labels, summary claims and search references Add migrations or a versioned document format, explicit validation, stable IDs and a visible import-error report. Preserve this rule: Action owners and decisions require transcript evidence; diarization labels do not identify a real person without user confirmation.
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. Save a job manifest with input hash, parameters and state. Write to temporary outputs, then atomically finalize only successful results; resume unfinished jobs without replacing originals.
4. Phase 4 — Permissions and integration failure. Bound file sizes and processing time, reject path traversal, and use argument arrays for subprocesses. Treat imported text as data and redact confidential source content from logs. Request integration credentials and permissions only for the enabled feature; show a disconnected state instead of mock results.
5. Phase 5 — Portable handoff. Export sources, manifests and outputs with checksums. Keep failed-job diagnostics and allow retry into a new output path; restore the database and file directory together. Include setup, operating limits, fixture walkthrough and shutdown/restart instructions in the README.
6. Phase 6 — Acceptance scenarios. A quoted action opens its exact segment; failed transcription retains audio and an uncertain speaker remains labeled unknown. Repeat the workflow after restart and with a denied permission or unavailable dependency; show recoverable failure rather than a success placeholder.

Acceptance
A quoted action opens its exact segment; failed transcription retains audio and an uncertain speaker remains labeled unknown.
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: [modern-python](https://github.com/trailofbits/skills/blob/main/plugins/modern-python/skills/modern-python/SKILL.md) — Set up Python projects with pyproject.toml, dependency management, linting, typing and automated checks. Review its instructions and compatibility before use; it does not grant deployment, data-access or publication permission.
Optional external skill: [web-design-guidelines](https://github.com/vercel-labs/agent-skills/blob/main/skills/web-design-guidelines/SKILL.md) — Review web interfaces for accessibility, keyboard focus, forms, navigation and interaction quality. Review its instructions and compatibility before use; it does not grant deployment, data-access or publication permission.
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.
Project rule — data model: consented recordings, timestamped segments, tentative speaker labels, summary claims and search references
Project rule — preserve this invariant: Action owners and decisions require transcript evidence; diarization labels do not identify a real person without user confirmation.
Project rule — acceptance evidence: A quoted action opens its exact segment; failed transcription retains audio and an uncertain speaker remains labeled unknown.

## 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.