context

A context-management tool for saving, listing, resuming, and checking the state of coding sessions. A saved snapshot records the goal, decisions, current progress, open issues, and next steps.

In plain words
What is it for?
Use it to save a durable session summary, resume work from that summary, list saved snapshots, or check remaining context space.
Why use it?
It reduces the need to reconstruct a long task when a session is compressed, ends, or is resumed later.

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/lukasrepublic/agentic-foundry/context
Any agent
npx skills add lukasrepublic/agentic-foundry --skill context
Clone the repo
git clone --depth 1 https://github.com/lukasrepublic/agentic-foundry

Made for: Claude Code, Codex.

Per session 72 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,230 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. 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 $0.00072 $0.01230
Opus 5 $0.00036 $0.00615
Sonnet 5 $0.00014 $0.00246
Haiku 4.5 $0.00007 $0.00123

Measured yesterday against content hash f5c4a0fc8d9c, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

context scanned grade A with 0 findings 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 yesterday.

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.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/context/SKILL.md · 83 lines

How it starts

The opening of the file, as written. The whole thing — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/foundry:context

The context lifecycle, an additive WRAP layer over native primitives. The four ctx-* skills fold into one /foundry:context <action> — each action is a THIN seam over a native primitive, realizing (extend, don't compete).

action native primitive it wraps foundry delta
snapshot /compact (lossy in-session) + --resume (verbatim replay) writes a DURABLE, hand-editable distilled arc-state file (§0–§6 schema) that survives session end + 30-day replay retention. Additive to --resume, not a replacement.
resume --resume <id> (verbatim) re-grounds a FRESH session from the distilled snapshot's §6 authoritative-artifact pointers (cheaper than verbatim replay; survives expiry).
list lists snapshots (the summary: line per file).
status the context sensor reports context budget / tier.

The snapshot schema (§0–§6)

The distilled arc-state file has seven short sections, numbered so resumers can cite them: §0 header (session id, date, mandate) · §1 the mandate/goal · §2 decisions taken (with why) · §3 current state (what is done/verified) · §4 open threads + blockers · §5 next steps · §6 authoritative artifact pointers (paths/commits/PRs a successor MUST re-read before trusting the prose — the load-bearing section).

Procedure

  • snapshot — distill the session arc into the §0–§6 schema file; §6 (authoritative artifact pointers) is load-bearing (a snapshot whose §6 is empty is a defect).
  • resume <id|path> — read the snapshot; re-read its §6 artifacts BEFORE trusting the distilled prose.
  • list / status — read-only.

Relationship to native (do NOT reinvent)

/compact (in-session lossy cost lever), --resume (verbatim replay), /rewind (within-session rollback) remain authoritative. This skill is the DURABLE, distilled, cross-session cold-start seam on top of them — additive only.

Mandate handoff

(feat-foundry-session-per-mandate, AC-SPM-3.) Session-per-mandate is the enforced advisory norm: one session per mandate, ended and handed off rather than compacted through. Prior operating-cost retrospective grounding showed this is the largest single cost lever — sessions that snapshot-and-respawn instead of compacting run compaction-free, at a fraction of the cache-read spend of a multi-mandate mega-session.

Read the full file on GitHub · 83 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. yesterday First seen · 83 lines · 72 tokens per session scan A f5c4a0fc8d9c

Subscribe to this mod's changes

context is a skill published in the GitHub repository lukasrepublic/agentic-foundry (1 stars, last pushed yesterday), licensed MIT. It adds 72 tokens to every session and 1,230 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.