session-loop

A set of tools for safely pausing and resuming multi-day projects worked on with coding agents. It records session endings, restarts, checkpoints, decisions, and checks for differences between project records and reality.

In plain words
What is it for?
Use it to wrap up a session, catch up on earlier work, save progress during a break, record decisions, or check whether a project ledger is still accurate.
Why use it?
It reduces lost context, repeated decisions, and unnoticed changes between work sessions.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/akshayrao14/git-practices/session-loop
Any agent
npx skills add akshayrao14/git-practices --skill session-loop
Clone the repo
git clone --depth 1 https://github.com/akshayrao14/git-practices

Made for: Claude Code, Codex.

Per session 241 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,771 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00241 $0.01771
Opus 5 $0.00120 $0.00886
Sonnet 5 $0.00048 $0.00354
Haiku 4.5 $0.00024 $0.00177

Measured 2d ago against content hash 4db68d1abd61, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

session-loop scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

The scan reads SKILL.md. This mod also ships 1 executable file (install.sh), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Enumerates other installed skillsmediumAgent snooping

Other skills' SKILL.md files reveal prompts, capabilities and secrets that should be invisible to peers.

Installs into the right skills dir for your agent (Codex `~/.codex/skills`, Claude Code `~/.claude/skills`, or open-standard `~/.agents/skills`). Restart your agent session afterward.
skills/session-loop/SKILL.md · 96 lines

How it starts

The opening of the file, as written. The whole thing — 96 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Session Loop

A pause/resume toolkit for multi-day projects worked on with agent assistants (Claude Code, Codex, etc.). Composes with append-only project ledgers — does not replace them.

Why this exists

Multi-day projects fail at the seams. The work itself is fine. The handoffs between sessions are where context evaporates, decisions get re-litigated, and silent drift compounds. This skill is the missing scaffolding around agent + project + ledger that makes pause/resume robust.

Sub-flows

Slash form When it fires Sub-skill file
/session-wrap End of session. User says "stop for today", "let's wrap up", "I'm done". sub/wrap.md
/session-catchup Start of session. User says "pick up where we left off", "what was I doing", "resume the migration". sub/catchup.md
/session-checkpoint Mid-session light save. User says "stepping away for an hour", "save my place". sub/checkpoint.md
/session-drift Sanity check world vs ledger. User says "check the world before I run helm", "is the ledger still accurate". sub/drift.md
/session-decide Log an ADR-style decision. User says "we just decided X, log it", "record this call". sub/decide.md
/session-open-loops Scan unresolved threads. User says "what's still open", "what did we defer". sub/open-loops.md
/session-compact-check Pre-compaction state persistence. User says "compact now", or context pressure is high at a task boundary. sub/compact-check.md

Shared artifact contract

These files live in the project root. Each sub-flow reads and/or writes a defined subset. The contract is convention — no sub-flow hard-fails if a file is missing, but they are most useful when all three exist.

File Lifecycle Owner sub-flows
NEXT_SESSION.md Ephemeral. Overwritten on every /session-wrap and /session-checkpoint. Consumed by /session-catchup. Write: wrap, checkpoint. Read: catchup, open-loops.
DECISIONS.md (or project's existing ledger, e.g. MIGRATION_LEDGER.md) Append-only. Never retro-edit; supersede with new dated entries. Write: decide. Read: catchup, drift, open-loops.
UPDATES.md Append-only narrative, newest-on-top, dated. Write: wrap. Read: catchup.

Read the full file on GitHub · 96 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 96 lines · 241 tokens per session scan B 4db68d1abd61

Subscribe to this mod's changes

session-loop is a skill published in the GitHub repository akshayrao14/git-practices (0 stars, last pushed 5d ago), licensed MIT. It adds 241 tokens to every session and 1,771 once invoked, about $0.0012 per session on Opus 5. A static security scan graded it B with 1 finding (enumerates other installed skills). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.