agents-context-management
installsforks
How to install agents-context-management?
paulrberg/agent-skills/agents-context-management
$npx -y skills add PaulRBerg/agent-skills --skill agents-context-managementCreate or polish repo agent context: README.md, AGENTS.md/CLAUDE.md, and installed project skills.
SKILL.md
| 1 | # Agents Context Management |
| 2 | |
| 3 | Create or polish repo-local context as one coherent system: human-facing README.md files, agent-facing AGENTS.md files |
| 4 | with companion CLAUDE.md symlinks, and existing project-installed skills under `.agents/skills`. |
| 5 | |
| 6 | Success means every selected target is grounded in repository evidence, respects its audience and scope, and passes the |
| 7 | narrowest repository-defined validation. Stop after reporting completed or planned changes, validation, and any |
| 8 | blockers. |
| 9 | |
| 10 | ## Model Optimization |
| 11 | |
| 12 | Optimize skills and other agent-facing context for GPT-5.6 and Claude Fable 5 while preserving README.md as clear |
| 13 | human-facing documentation. The summaries below are reminders, not substitutes for the live guides. Read both guides |
| 14 | before complex, long-running, multi-tool, or orchestration-heavy context work because their recommendations may evolve. |
| 15 | |
| 16 | - [GPT-5.6 prompting guidance](https://developers.openai.com/api/docs/guides/prompt-guidance-gpt-5p6): Prefer lean, |
| 17 | outcome-first prompts that specify the goal, success and stopping criteria, constraints, evidence, permission |
| 18 | boundaries, tool routing, output shape, and validation. Remove redundant scaffolding and evaluate changes on |
| 19 | representative tasks. |
| 20 | - [Claude Fable 5 prompting guidance](https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/prompting-claude-fable-5): |
| 21 | Use concise instructions that explain intent and boundaries; avoid over-prescription and scope creep; tune effort |
| 22 | deliberately; ground progress claims in tool evidence; and make long-run verification and scaffolding explicit when |
| 23 | needed. |
| 24 | |
| 25 | ## Choose a Workflow |
| 26 | |
| 27 | Choose exactly one workflow and read only its reference. |
| 28 | |
| 29 | | User intent | Workflow | Reference | |
| 30 | | --------------------------------------------------------------- | ---------------------------- | ----------------------------------------- | |
| 31 | | Update, refresh, sync, prune, polish, repair, or fix context | `polish` | `references/brain-polish.md` | |
| 32 | | Create, initialize, generate, or regenerate context files | `create` | `references/create-docs.md` | |
| 33 | | Audit, check, review, inspect, or suggest changes without edits | `polish` in `--dry-run` mode | `references/brain-polish.md` | |
| 34 | | Create or scaffold a skill | Stop | Refer to `skills/create-skill` | |
| 35 | | Install, discover, remove, or rename a skill | Stop | Use a dedicated skill-management workflow | |
| 36 | |
| 37 | If the intent is unclear, select `polish` in `--dry-run` mode and report the smallest useful planned change set. |
| 38 | |
| 39 | ## Authority |
| 40 | |
| 41 | - Explicit create, update, polish, repair, fix, or equivalent intent authorizes in-scope local writes. Inspection-only |
| 42 | intent and `--dry-run` do not. |
| 43 | - Require explicit confirmation before deleting README.md, AGENTS.md, or CLAUDE.md entries. `--force` authorizes |
| 44 | documented overwrites, not deletions. |
| 45 | - Treat a broad write request as authorization for the requested scope. Otherwise, preview a change set larger than a |
| 46 | handful of files and stop before writing. |
| 47 | - Do not expand from documentation work into source changes, skill creation, or external writes. |
| 48 | |
| 49 | ## Arguments |
| 50 | |
| 51 | - `path`: Optional repo-relative subtree. Restrict documentation, package-root, and project-skill discovery to that |
| 52 | subtree. |
| 53 | - `skill-name ...`: Optional filters for existing `.agents/skills/<name>/` targets during `polish`. |
| 54 | - `--root-only`: Select only root README.md, AGENTS.md, and CLAUDE.md targets. Exclude project skills unless explicitly |
| 55 | selected by `skill-name`. |
| 56 | - `--dry-run`: Report planned writes and concise diffs without changing files. |
| 57 | - `--preserve`: During `polish`, keep accurate user-authored prose and structure; fix only drift and obvious noise. |
| 58 | - `--minimal`: Produce the smallest context that still meets the completion bar. |
| 59 | - `--thorough` / `--full`: Perform deeper analysis only where it adds durable, repository-specific context. |
| 60 | - `--force`: During `create`, regenerate existing README.md or AGENTS.md targets without prompting. Never applies to |
| 61 | skills or deletions. |
| 62 | |
| 63 | If `--minimal` and `--thorough` / `--full` are both present, make no writes and ask the user to choose. Report |
| 64 | unrecognized flags; continue only when they cannot change scope, safety, or write behavior. |
| 65 | |
| 66 | ## Repository Guard Rail |
| 67 | |
| 68 | Run before discovery or writes: |
| 69 | |
| 70 | ```sh |
| 71 | cwd="$(pwd -P)" |
| 72 | case "$cwd" in |
| 73 | /) printf 'abort: refusing to run at the filesystem |
People who liked this also viewed
6 picksType
- find-skillsHelps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skSkillsJul 2026
- grill-meA relentless interview to sharpen a plan or design.SkillsJul 2026
- vercel-react-best-practicesVercel's official collection of agent skillsSkillsJul 2026
- improve-codebase-architectureScan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.SkillsJul 2026
- setup-matt-pocock-skillsConfigure this repo for the engineering skills — set up its issue tracker, triage label vocabulary, and domain doc layout. Run once before first use of the other engineering skills.SkillsJul 2026
- triageMove issues and external PRs through a state machine of triage roles — categorise, verify, grill if needed, and write agent-ready briefs.SkillsJul 2026
