Skip to content

CHANGELOG.md

Latest commit

 

History

History
102 lines (76 loc) · 6.42 KB

CHANGELOG.md

File metadata and controls

102 lines (76 loc) · 6.42 KB

Changelog

All notable changes to agentlayer are documented here.

[Unreleased]

[0.6.0] - 2026-04-23

Added

  • agentlayer init — recommended single-entrypoint command. Detects repository state (fresh vs. already has AI files), asks once which runtime(s) to use, prints a preflight summary (target repo, project type, existing AI files, chosen action, runtimes), confirms, and routes to install or standardize. Path argument is optional; defaults to the current directory. audit, install, and standardize remain available for advanced/scripted use.
  • README recut around the first-use path (agentlayer init). Internal structural detail moved to MANUAL.md.
  • README "Alternative — let your AI install it" section now states the prerequisites explicitly: the assistant must have local read/write access to the repository; cloud-only chat tools do not work.

Changed (UX, breaking in edge cases)

  • agentlayer install now refuses to run on a repository that already has any AI-related file (.ai/, AGENTS.md, CLAUDE.md, .github/copilot-instructions.md, .github/instructions/, .cursor/rules/agentlayer.mdc, or AGENTLAYER.md). The CLI exits non-zero and points at agentlayer standardize. The previous behavior (warn + optional prompt to continue) could leave a repo in a mixed state.
  • agentlayer install --non-interactive without --runtimes now exits non-zero with an explicit error. Previously it silently fell back to a detected default (e.g. codex,claude,generic) and generated adapters the user had not asked for.
  • install and standardize post-run output replaced the prose guidance with visually delimited, copy-paste-ready prompt blocks — one per selected runtime, with wording aligned to MANUAL.md "What To Ask The AI After Install".

Fixed

  • agentlayer install . no longer renders Project Name | . in generated docs. The target path is normalized with cd && pwd -P before any path operation.
  • install.sh now emits a non-fatal warning when run from a volatile path (/tmp, /var/folders, /var/tmp). Clones into those locations produced a symlink that broke on next OS cleanup.
  • README CLI-install example now clones into ~/.agentlayer instead of /tmp/agentlayer to avoid the same footgun.

[0.5.0] - 2026-04-17

Changed

  • Project renamed from AI-Ready Bootstrap to agentlayer.
  • GitHub repository moved to iMark21/agentlayer.
  • CLI command renamed from ai-ready to agentlayer.

[0.4.0] - 2026-04-17

Changed

  • assistant-installer/SKILL.md removed. PROMPT.md is now the single universal installer — it works identically with Codex, Claude Code, Copilot, Cursor, or any other AI, and is self-contained with all adapter formats embedded.
  • assistant-installer/addon/ removed. The Codex skill add-on was a convenience wrapper for an install that runs once per repo. The prompt is sufficient.
  • assistant-installer/README.md simplified from a routing table to a plain explanation of PROMPT.md.
  • README and MANUAL.md updated to remove all references to SKILL.md and the addon.

[0.3.3] - 2026-04-17

Added

  • PROMPT.md and SKILL.md now end with concrete copy-paste-ready examples using real project context. After install the AI shows the user three or more prompts they can run immediately — using real module names and real opportunities noticed during the audit, not generic placeholders.
  • "Done means" checklist in SKILL.md updated: install is not complete until the user has actionable next steps.

[0.3.2] - 2026-04-16

Fixed

  • PROMPT.md and SKILL.md now require the AI to ask which runtime(s) the user wants BEFORE creating any files. Previously the [auto] default caused the AI to install all runtimes and search the web for Copilot and Cursor adapter formats.
  • All adapter format templates are now embedded inline in PROMPT.md and SKILL.md. The AI no longer needs to search the web to know the correct format for CLAUDE.md, AGENTS.md, copilot-instructions.md, .mdc, or AGENTLAYER.md.
  • Added explicit "Do not search the web" rule to both files.
  • Synced addon/agentlayer-installer/SKILL.md with parent SKILL.md.

[0.3.1] - 2026-04-16

Changed

  • README install section collapsed from four per-AI blocks into two options: "Let your AI do it (any AI)" and "Use the CLI". The AI-assisted path is now a single universal prompt that works with Codex, Claude Code, Copilot, Cursor, or any other AI — the spec is fetched from PROMPT.md and the workflow asks which runtimes to use.

[0.2.1] - 2026-04-16

Changed

  • README rewritten as a scenario-driven guide: "Does this sound familiar?", agent flow diagram, copy-paste install blocks per runtime (Claude Code, Codex, Copilot, Cursor, generic), and three real workflow examples (feature, bug fix, investigation)
  • Install section reorganised by AI tool instead of Path A / Path B — each runtime has a self-contained block with the exact prompt or command to use

[0.2.0] - 2026-04-16

Added

  • Compelling README with Before/After table, Quick Start, and "Who is this for" section
  • MANUAL.md Quick Start section at the top
  • Closing section in MANUAL.md summarizing the five core pieces of the system
  • CHANGELOG.md
  • CODE_OF_CONDUCT.md
  • SECURITY.md
  • GitHub issue and PR templates

Changed

  • README restructured as a public-facing landing page
  • MANUAL.md examples now use generic placeholder names instead of personal git identity
  • .gitignore expanded with common editor and OS patterns
  • assistant-installer/README.md now includes a context paragraph before the routing table
  • CONTRIBUTING.md updated with external contributor welcome

Removed

  • "Private repo" reference from README

[0.1.1] - 2026-04-16

Added

  • Explicit Codex skill-install path via assistant-installer/addon/README.md
  • Ready-to-copy add-on package at assistant-installer/addon/agentlayer-installer/

[0.1.0] - 2026-04-16

Added

  • CLI (bin/agentlayer) with audit, install, and standardize commands
  • Multi-runtime selection: Codex, Claude Code, Copilot, Cursor, generic
  • Project type detection: Android, iOS, web, backend, generic
  • Canonical .ai/ layer with context, rules, agents, and skills
  • AI-assisted install path (assistant-installer/SKILL.md, PROMPT.md)
  • First-pass context bootstrap workflow (proj-context-bootstrap)
  • Git governance with optional pre-commit hook and identity configuration
  • GitHub Actions CI with syntax checks and smoke tests for Android, iOS, and standardize modes
  • install.sh for external CLI adoption