codecall v1.3.0

Every session now ends in a one-page receipt (./codecall/NNNN-.html) with the score "Could explain N of M" and a streak line. SKILL.md trimmed and re-led with the slot; one-line install: npx skills add GAURAV-1313/codecall. Test pins fixed.

codecall v1.2.2

Small patch release

• README now links the landing page, npm, and GitHub at the top.
• Lockfile refresh (no dependency changes, just pruned stale optional-platform variants).

No behavior changes for any install path.

codecall v1.2.1

Small patch release

• README: added a second worked example (retry/backoff logic) alongside the existing auth example, so it's clear codecall isn't just for security-flavored code.
• Fixed a high-severity dev-dependency advisory (nanoid, GHSA-2v37-7h3g-55p8) — dev-only, never shipped to users, but fixed for repo hygiene.
• Removed a stray self-referential codecall dependency that had crept into package.json/package-lock.json from a local install in this directory — never actually published, but worth cleaning up.

No behavior changes for any install path.

codecall v1.2.0

Deterministic auto-invocation + context-window relief

Two additions on top of v1.1.0's dual-plugin release:

Stop hook (both plugin installs, default-on) — a small local script now deterministically re-checks trigger-policy.md before the agent's final response, instead of relying only on the model remembering to read AGENTS.md/CLAUDE.md. It never judges recommend/optional/skip itself (that stays a model decision) — it only forces the check to happen, with a cheap pre-filter (ignores doc/test/lockfile-only changes) and per-project dedupe so it doesn't repeat itself. Claude Code gets a silent nudge; Codex's confirmed mechanism forces one extra turn instead.

codecall-planner subagent (Claude Code plugin only) — the read-heavy context-gathering and planning phase now runs in an isolated subagent instead of the main conversation, so it no longer permanently occupies your main context window. The interactive teach/confidence/quiz loop still runs in the main thread, since a subagent can't pause mid-run for your answers. Codex has no reliable way for a skill to invoke one specific named subagent, so it keeps planning in-context for now.

Also: SKILL.md's frontmatter description trimmed from 488 to 188 characters (lower per-invocation token cost), and the README rewritten for a non-technical audience with a concrete example of what a session looks like.

What changes for existing users

• npm standalone (npm install -g codecall): no behavior change — only the shorter description ships. Neither the hook nor the subagent can attach to this install path (no plugin identity). Run npm update -g codecall to pick it up.
• Claude Code plugin: gets both the hook and the planner subagent. Update with /plugin marketplace update codecall then /reload-plugins. You may see a one-time trust/review prompt for the new bundled hook.
• Codex plugin: gets the hook only (more intrusive nudge — forces an extra turn rather than a silent one). Update with codex plugin marketplace upgrade codecall and reinstall codecall@codecall if needed.

Install

codex plugin marketplace add https://github.com/GAURAV-1313/codecall
codex plugin add codecall@codecall

/plugin marketplace add https://github.com/GAURAV-1313/codecall
/plugin install codecall@codecall
/reload-plugins

npm install -g codecall

codecall v1.1.0

Codex and Claude Code marketplace plugins

Codecall now ships as a plugin for both Codex and Claude Code, backed by one shared skill source (plugins/codecall/), alongside the existing npm standalone installer.

Install

Codex plugin

codex plugin marketplace add https://github.com/GAURAV-1313/codecall
codex plugin add codecall@codecall

Claude Code plugin

/plugin marketplace add https://github.com/GAURAV-1313/codecall
/plugin install codecall@codecall
/reload-plugins

Invoke with /codecall:codecall.

npm standalone

npm install -g codecall

Invoke with $codecall (Codex) or /codecall (Claude Code).

No API key, no external learning service, no automatic teaching without consent.

codecall v1.0.0

Renames the product, npm package, CLI, Codex skill, Claude Code skill, templates, and public API from learn to codecall.

learn v0.1.6

Adds an explainable, selective automatic learning recommendation policy with strong/moderate signals, exclusions, evidence-backed decisions, and session-local duplicate suppression.

learn v0.1.5

Adds Claude Code compatibility. One installation now places the shared Agent Skills-standard learn skill in both Codex and Claude Code personal skill directories.

learn v0.1.4

Adds implementation-grounded points to remember and cross-project edge-case guidance to learning summaries. The README now uses the verified GitHub source-tarball installer.

learn v0.1.2

Adds implementation-grounded points to remember and cross-project edge-case guidance to the final learning summary.