Skip to main content
Glama

_delimit_context_impl

Manage a venture-scoped, versioned context filesystem to persist plans, decisions, and artifacts across AI sessions. Perform six actions: init, read, write, list, snapshot, and branch.

Instructions

Unified context-filesystem entry point — dispatches to one of six actions.

Manages a venture-scoped, versioned context filesystem under ~/.delimit/context// so plans, decisions, and artifacts survive across sessions and across models. This is the cross-model- continuity store: write once, read from any later session or any other assistant.

When to use: as the single MCP-registered context surface (delimit_context) when the caller wants to pick the action by name in one call rather than choosing a specific delimit_context_* alias. When NOT to use: from internal code paths — prefer the specific alias (delimit_context_read, delimit_context_write, delimit_context_snapshot, etc.) so each action's docstring, args, and side-effect notes show up at the right call site. For ephemeral, conversation-scoped memory use delimit_memory_store / delimit_memory_search instead — those are NOT venture-namespaced or versioned.

Sibling contrast: each delimit_context_ wrapper below is a thin alias over this implementation; they exist so the action's docstring lives at the right name. This is the dispatch core. The context FS is venture-scoped and versioned (snapshot/branch); delimit_memory_* is conversation-scoped and unversioned. Snapshot vs branch: snapshot is an immutable point-in-time copy (history/ rollback), branch is a mutable write-isolated fork that can be merged back into main. Neither touches git or any code repository.

Side effects: all six actions are free-tier (no require_premium gate in this dispatcher). Each routes to a distinct context-FS backend function and is wrapped via _with_next_steps for orchestrator hints. Per action:

  • "list" — read-only enumeration of /artifacts/*. Returns [] (no error) if the venture or artifacts dir does not exist.

  • "read" — read-only load of one artifact. Returns {"error": ...} if the named artifact is absent.

  • "init" — WRITES. Creates the venture directory, the memory/plans/artifacts/snapshots/branches subdirs, and manifest.json if absent. Idempotent.

  • "write" — WRITES/overwrites /artifacts/.json and bumps the manifest version counter. Overwrites silently if the artifact name already exists.

  • "snapshot" — WRITES. Copies the venture's artifacts/ and memory/ into a timestamped (optionally labeled) snapshot dir plus a snapshot manifest. Does NOT bump the version counter.

  • "branch" — depends on branch_action. "list" is read-only. "create" WRITES a new branch fork (copy of artifacts/ + memory/) and errors if the branch already exists. "merge" MUTATES the venture's main artifacts/ and memory/ with the branch's files, then DELETES the branch dir and bumps the version counter; errors if the branch is not found. Errors are deterministic ({"error": "..."}): an unknown top-level action, an unknown branch_action, or a missing branch_name on create/merge all short-circuit before the backend call.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
actionNoWhich context operation to perform. One of "init", "read", "write", "list", "snapshot", "branch". Default "list". Other values return a deterministic error.list
ventureNoVenture/project namespace key — selects the ~/.delimit/context/<venture>/ tree. Used by every action. Default "default".default
nameNoArtifact name, used as the <name>.json file key. Required for action="read" and action="write". Ignored by other actions.
contentNoArtifact text body. Used only when action="write".
artifact_typeNoType hint stored on the artifact — "text", "json", "code", or "plan". Used only when action="write". Default "text". Affects the stored type hint, not the storage format.text
labelNoOptional human-readable snapshot label, appended to the timestamp in the snapshot dir name. Used only when action="snapshot".
branch_actionNoBranch sub-action — "list", "create", or "merge". Used only when action="branch". Default "list".list
branch_nameNoBranch name. Required when action="branch" with branch_action="create" or "merge"; ignored for "list".

Output Schema

TableJSON Schema
NameRequiredDescriptionDefault

No arguments

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations provided, so description fully covers side effects per action: read-only vs write ops, idempotency, silent overwrite, deterministic errors, and destructive merge behavior (deletes branch).

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Well-structured with sections, but could be slightly more concise; the side-effect per-action list is thorough but verbose.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given 8 parameters, no annotations, and complex dispatch logic, the description is comprehensive, covering all actions, errors, and comparisons with sibling/memory tools. Output schema exists, so return details are not needed.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, but the description adds practical context like conditional dependencies (branch_action requires branch_name) and default behaviors, beyond the schema's type/default fields.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states it is a 'unified context-filesystem entry point' that dispatches to six actions, and distinguishes itself from sibling aliases like delimit_context_read, delimit_context_write, etc., by being the dispatch core.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Explicit 'When to use' and 'When NOT to use' sections, compares to memory tools, and suggests preferring specific aliases for internal code paths.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Install Server

Other Tools

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/delimit-ai/delimit-mcp-server'

If you have feedback or need assistance with the MCP directory API, please join our Discord server