delimit

Official

Overview Schema Source Related Servers Discussions Admin

Delimit Soul Capture

delimit_soul_capture

Capture session state as a soul so any model can resume work exactly where it stopped. Save active task, decisions, next steps, and context before context fills or session ends.

Instructions

Capture session state as a 'soul' for cross-model resurrection.

When to use: at session end or when context gets full, to save what you're working on so the next session in any model can pick up where you left off. When NOT to use: for general memory writes (use delimit_memory_store) or full handoff orchestration (delimit_session_handoff).

Sibling contrast: delimit_session_handoff writes a structured handoff for the next session; this writes a state soul with active-task pointers for delimit_revive. Session Phoenix, when installed, also captures git and transcript state.

Side effects: writes a local soul record through Session Phoenix when it is installed, or through the shipped Free session_continuity core. Explicit local capture/handoff/revive is Free; automatic transcript salvage remains a Pro Session Phoenix capability. The current model is detected in both tiers; git state is collected only when Session Phoenix is installed. Plain strings become one entry (with embedded commas preserved); native/JSON lists carry multiple entries.

Input Schema

TableJSON Schema

Name	Required	Description	Default
`blockers`	No	Blockers as one string, a native list, or a JSON list string.
`decisions`	No	Key decisions as one string, a native list, or a JSON list string.
`next_steps`	No	Next steps as one string, a native list, or a JSON list string.
`active_task`	No	What you're currently working on (one line).
`key_context`	No	Important context as one string, a native list, or a JSON list string.
`task_status`	No	One of "in_progress", "blocked", "almost_done".	in_progress
`tokens_used`	No	Estimated tokens consumed this session.
`project_path`	No	Project path to capture the soul for. Empty = auto-detect from cwd.
`context_fullness`	No	0.0-1.0 representing context-window fullness.

Output Schema

TableJSON Schema

Name	Required	Description	Default
No arguments

Schema Changelog

Changes observed during successful MCP inspections.

Changed12 schema fields changedv4.18.8Sep 8, 2026

addedInput schema / properties / blockers / anyOf

Added value: +[
+  {
+    "type": "string"
+  },
+  {
+    "items": {
+      "type": "string"
+    },
+    "type": "array"
+  },
+  {
+    "type": "null"
+  }
+]

changedInput schema / properties / blockers / description

Previous value: -"Comma-separated blockers."New value: +"Blockers as one string, a native list, or a JSON list string."

removedInput schema / properties / blockers / type
```
Removed value: -"string"
```

addedInput schema / properties / decisions / anyOf

Added value: +[
+  {
+    "type": "string"
+  },
+  {
+    "items": {
+      "type": "string"
+    },
+    "type": "array"
+  },
+  {
+    "type": "null"
+  }
+]

changedInput schema / properties / decisions / description

Previous value: -"Comma-separated key decisions made this session."New value: +"Key decisions as one string, a native list, or a JSON list string."

removedInput schema / properties / decisions / type
```
Removed value: -"string"
```

addedInput schema / properties / key_context / anyOf

Added value: +[
+  {
+    "type": "string"
+  },
+  {
+    "items": {
+      "type": "string"
+    },
+    "type": "array"
+  },
+  {
+    "type": "null"
+  }
+]

changedInput schema / properties / key_context / description

Previous value: -"Comma-separated important context for next session."New value: +"Important context as one string, a native list, or a JSON list string."

removedInput schema / properties / key_context / type
```
Removed value: -"string"
```

addedInput schema / properties / next_steps / anyOf

Added value: +[
+  {
+    "type": "string"
+  },
+  {
+    "items": {
+      "type": "string"
+    },
+    "type": "array"
+  },
+  {
+    "type": "null"
+  }
+]

changedInput schema / properties / next_steps / description

Previous value: -"Comma-separated next steps."New value: +"Next steps as one string, a native list, or a JSON list string."

removedInput schema / properties / next_steps / type
```
Removed value: -"string"
```

Changed1 schema field changedv4.13.2Sep 5, 2026

addedInput schema / properties / project_path

Added value: +{
+  "default": "",
+  "description": "Project path to capture the soul for. Empty = auto-detect from cwd.",
+  "type": "string"
+}

Changed2 schema fields changedv4.7.9Jun 9, 2026

changedInput schema / properties / context_fullness / description

Previous value: -"0.0-1.0 how full the context window is."New value: +"0.0-1.0 representing context-window fullness."

changedInput schema / properties / task_status / description

Previous value: -"in_progress, blocked, or almost_done."New value: +"One of \"in_progress\", \"blocked\", \"almost_done\"."

Addedv4.5.5May 7, 2026

TDQS

A4.5/5.0

Behavior4/5

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

With only minimal annotations (readOnlyHint=false, destructiveHint=false), the description carries the behavioral burden. It discloses side effects ("writes a local soul record"), tier-dependent behavior (Free vs Pro Session Phoenix, git state collection only when Phoenix is installed), and edge-case handling for list parameters (plain strings vs native/JSON lists). It does not mention overwrite/idempotency semantics, but the provided context is substantial and consistent with annotations.

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?

The description is longer than average but well-structured with clear sections: purpose, when to use, when not to use, sibling contrast, and side effects. It front-loads the core purpose and usage guidance. Some details (Free/Pro breakdown, git collection conditions) add length but are relevant to behavioral transparency and invocation context, so the length is justified.

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

Completeness4/5

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

Given the tool's complexity (9 parameters, many optional, output schema present), the description covers the essential invocation context: when to call, which sibling to use instead, side effects, and parameter value semantics. It does not need to explain return values because an output schema exists. It is complete enough for an agent to select and invoke the tool correctly.

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%, so the baseline is 3. The description adds meaningful behavior beyond the schema by explaining that "Plain strings become one entry (with embedded commas preserved); native/JSON lists carry multiple entries." This directly informs how to pass parameters like blockers, decisions, and key_context, which the schema only describes structurally.

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 opens with a specific verb and resource: "Capture session state as a 'soul' for cross-model resurrection." It further distinguishes itself from delimit_session_handoff by stating that this tool writes "a state soul with active-task pointers for delimit_revive," while the handoff writes a structured handoff. This makes the tool's role unambiguous even among a large sibling set.

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?

The description explicitly tells the agent when to use the tool ("at session end or when context gets full") and when not to use it: "for general memory writes (use delimit_memory_store) or full handoff orchestration (delimit_session_handoff)." It also contrasts with the sibling tool in detail, leaving no ambiguity about selection.

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

Deploy Server