Docs / Features
The product surface, without the icon wall.
Snipara is easier to understand when grouped by what it adds to a run: grounded retrieval, reviewed memory, optional execution, and operational visibility.
High-level view
- Code Context keeps workspace-wide engineering rules separate from project files, shared collections, project scopes, and health.
- Repository continuity is a daily workflow: Start Work Brief, What Changed For Me, and PR Answer Packs.
- snipara_context_query stays the default path for docs and narrative context.
- Code structure belongs to the code graph and the snipara_code_* tools.
- Context authority keeps source_url, commit_sha, freshness, confidence, validation state, and stale warnings visible.
- Team Sync turns parallel repository work into a shared brief with advisory Team evidence requests and enforceable Enterprise proof gates.
- Snipara Orchestrator is the Enterprise control plane for proof gates, drift checks, htasks, and explicit multi-agent coordination.
- PR Answer Packs turn GitHub pull requests into scoped context artifacts for Claude Code, GitHub Copilot, Codex, Cursor, and other AI coding clients.
- Supported PDF, DOCX, PPTX, SVG, and VSDX files now enter a deterministic binary parser lane instead of being ignored.
- Helvabase owns RFP, proposal, and client dossier workflows while Snipara stays focused on developer context.
- Shared context is project-aware, so one team can keep different rules and reference libraries per project.
- Full automation is optional; the simple path is still hosted MCP plus reviewed memory.
Block 01 / Core retrieval
Ground the run before the model improvises.
Use Code Context for reusable engineering rules, then project-scoped retrieval for the repository, internal, code, or research work in front of the model.
Block 02 / Sandbox and execution
Add execution only when the task needs it.
Snipara can stay retrieval-only, or extend into Snipara Sandbox, proof gates, and workflow execution when validation, automation, or multi-step work is worth the extra layer.
Block 03 / Code and operations
Make repository context visible inside pull requests.
Use the code graph for structural questions, let GitHub sync route docs, code, and supported binaries into the right lane, then publish PR Answer Packs and Team Sync briefs so agents start from scoped repository cognition.