Agent Plugins Marketplace
← All plugins

instruction-placement

v0.15.11

Routes agent-instruction content to the surface that loads it at the right moment. The audit skill sweeps a repository's instruction layer and its ordinary markdown for content whose scope is narrower than the surface carrying it, meaning conventions keyed to one file type or one subtree sitting in an always-loaded CLAUDE.md or AGENTS.md, and for normative conventions stranded in documentation Claude never loads at all, then classifies each against a routing rubric and proposes a destination whose `paths:` glob is machine-validated before it is ever offered. Safety-class content (irreversible actions, secrets, data integrity, external publication, compliance, agent authority) is hard-denied from demotion and reported as held back rather than proposed, because demotion trades guaranteed presence for conditional presence: a deferred surface is absent until a read matches it, absent after a compaction until that trigger recurs, and never inherited by a subagent, which re-acquires it only by reading a covered path itself. Every accepted move regenerates an always-loaded index of deferred surfaces, which is what keeps a demoted rule discoverable from any context that has not happened to touch a path it covers. The audit is read-only and emits a diffable findings artifact; realignment is a separate skill gated per item with no blanket-approve path; a deterministic check skill gates that every rule glob still resolves and the index is current; a migrate skill moves a repository to AGENTS.md as the content home and keeps a CLAUDE.md shim while one is needed; and a setup skill verifies the one thing no other gate can see: that nothing in the repository stops Claude Code reading the index target, since a CLAUDE.md in the working directory or above it is read instead of the AGENTS.md beside it.

Claude Code6 Skills

By Melodic SoftwareLicense: MIT20 GitHub starsUpdated 59 minutes ago

Directory evidence

Runtimes
Claude Code
Parsed components
6 skill or MCP entries
Source updated
Sep 28, 2026
Manifest status
Canonical path parsed

The directory validates manifest shape and source location. It does not execute the plugin or provide a security endorsement. Review the indexing methodology →

Install instruction-placement for Claude Code

Installs for the current user
claude plugin marketplace add IchenDEV/agent-plugin-mkt
claude plugin marketplace update agent-plugin-marketplace
claude plugin install instruction-placement@agent-plugin-marketplace

Paste and run these commands in a terminal with Claude Code. They add and refresh the PluginsMP catalog, then install this plugin.

The installer fetches third-party code from the source repository shown on this page. This directory validates manifest structure and source location, but does not perform a security audit; review the manifest, components, and source before installing.

Get the source manually
git clone https://github.com/melodic-software/claude-code-plugins

Clone the source repository, then follow its setup instructions to add the plugin to a compatible client. The plugin root is plugins/instruction-placement/.

Plugin files

plugins/instruction-placement/
├── .claude-plugin/plugin.json
├── skills/audit/SKILL.md
├── skills/check/SKILL.md
├── skills/delta/SKILL.md
├── skills/migrate/SKILL.md
├── skills/realign/SKILL.md
└── skills/setup/SKILL.md

Included Skills6

auditskills/audit/SKILL.md

Read-only sweep for instruction content on the wrong surface: demote always-loaded conventions that belong on one file type or subtree; promote stranded docs conventions. Safety rails are hard-denied from demotion. Use when: 'my CLAUDE.md is too long', 'convert this to rules', 'what should be a path-scoped rule', 'move conventions to .claude/rules', 'find conventions in our docs', 'nested CLAUDE.md candidates', 'audit instruction placement'. Sibling realign applies accepted findings.

checkskills/check/SKILL.md

Pass/fail gate over path-scoped instruction surfaces: every .claude/rules/ glob must resolve, and the always-loaded rules index must match disk. Failures are SILENT in Claude Code. Use when: 'check my rules', 'are my path-scoped rules actually firing', 'is the rules index stale', 'validate paths frontmatter', 'why is my rule not loading', 'CI gate for .claude/rules', or after a change to a rule's paths: or the rules tree. Read-only; sibling realign applies fixes.

deltaskills/delta/SKILL.md

Re-run the placement audit and report only what MOVED since the last run, above a noise budget. Declined findings stay declined. Use when: 'what changed since the last placement audit', 'placement delta', 're-run the instruction-placement audit', 'anything new to move', 'did any rule glob break', 'weekly instruction-placement check', or from a scheduled lane. Read-only. Realign still owns every change.

migrateskills/migrate/SKILL.md

Move instruction content to AGENTS.md as the one content home, keeping a one-line CLAUDE.md shim while a shim is what makes it load. Plans first; every write is operator-gated. Use when: 'migrate to AGENTS.md', 'plan the AGENTS.md migration', 'move CLAUDE.md content to AGENTS.md', 'add the AGENTS.md shim', 'our CLAUDE.md should be one line', 'share instructions with Codex and Cursor', 'can we drop the CLAUDE.md shims yet'. Not the placement sweep (audit) or applying findings (realign).

realignskills/realign/SKILL.md

Apply an instruction-placement audit's findings behind an explicit per-item human gate. Consumes the audit artifact; never re-judges. Each accepted finding is the whole move. Hard-denied safety content has no path here. Use when: 'apply the placement findings', 'do the migration', 'move those conventions to rules', 'execute finding IP-004', 'realign our instruction layer', 'the audit says move it, do it'. No blanket-approve. Audit proposes; this applies.

setupskills/setup/SKILL.md

Verify instruction-placement's prerequisites and resolve its effective configuration for this repository. Confirm that the index target exists AND that nothing in the repository stops Claude Code loading it (a root CLAUDE.md is read instead of the AGENTS.md beside it, so an unimported AGENTS.md index is inert while every other gate reports green), that `git` backs tracked-file discovery, and that the Claude Code CLI plus `jq` are present for the optional empirical load probe. Reports the resolved index target, breadth ceiling, and index row cap, naming which came from configuration and which from a default. Use when: 'set up instruction-placement', 'configure instruction-placement', 'where will the index go', 'why is my index not loading', 'is instruction-placement working', or before a first audit on a new repository. Actions: check (read-only verification, default) | apply (point at each remediation; writes nothing on its own). Re-runnable and safe.

Plugin manifests1

plugins/instruction-placement/.claude-plugin/plugin.json
{
  "$schema": "https://json.schemastore.org/claude-code-plugin-manifest.json",
  "name": "instruction-placement",
  "version": "0.15.11",
  "description": "Routes agent-instruction content to the surface that loads it at the right moment. The audit skill sweeps a repository's instruction layer and its ordinary markdown for content whose scope is narrower than the surface carrying it, meaning conventions keyed to one file type or one subtree sitting in an always-loaded CLAUDE.md or AGENTS.md, and for normative conventions stranded in documentation Claude never loads at all, then classifies each against a routing rubric and proposes a destination whose `paths:` glob is machine-validated before it is ever offered. Safety-class content (irreversible actions, secrets, data integrity, external publication, compliance, agent authority) is hard-denied from demotion and reported as held back rather than proposed, because demotion trades guaranteed presence for conditional presence: a deferred surface is absent until a read matches it, absent after a compaction until that trigger recurs, and never inherited by a subagent, which re-acquires it only by reading a covered path itself. Every accepted move regenerates an always-loaded index of deferred surfaces, which is what keeps a demoted rule discoverable from any context that has not happened to touch a path it covers. The audit is read-only and emits a diffable findings artifact; realignment is a separate skill gated per item with no blanket-approve path; a deterministic check skill gates that every rule glob still resolves and the index is current; a migrate skill moves a repository to AGENTS.md as the content home and keeps a CLAUDE.md shim while one is needed; and a setup skill verifies the one thing no other gate can see: that nothing in the repository stops Claude Code reading the index target, since a CLAUDE.md in the working directory or above it is read instead of the AGENTS.md beside it.",
  "author": {
    "name": "Melodic Software",
    "email": "[email protected]"
  },
  "license": "MIT",
  "userConfig": {
    "index_drift_hook_enabled": {
      "type": "boolean",
      "title": "index drift hook",
      "description": "PostToolUse notice when a write inside a .claude/rules tree leaves the generated index stale. Advisory and non-blocking; the authoritative gate is /instruction-placement:check in CI. Costs a string comparison on writes outside a rules tree.",
      "default": true
    },
    "breadth_max": {
      "type": "number",
      "title": "over-broad breadth ceiling (percent)",
      "description": "Percent of tracked files above which a rule's paths: glob is reported over-broad. Advisory only, never fails the check gate. Raise it in a repository where one extension legitimately covers most files.",
      "default": 75
    },
    "index_max_rows": {
      "type": "number",
      "title": "index rows before grouping",
      "description": "Surfaces listed individually in the generated index before the remainder is grouped by directory with a count. The index is always-loaded, so this bounds its own cost.",
      "default": 40
    }
  },
  "keywords": [
    "instruction-placement",
    "claude-md",
    "agents-md",
    "claude-rules",
    "path-scoped-rules",
    "progressive-disclosure",
    "context-economy",
    "audit",
    "realign",
    "skill"
  ]
}

If you maintain this plugin, link to this source-backed listing from your README so users can review its manifest and indexed components.

[instruction-placement on Agent Plugins Marketplace](https://pluginsmp.com/plugins/instruction-placement)