sage-artifact-retention
Preserve high-value Sage contracts, handoffs, reviews, transcripts, and evidence in a private project-local history while keeping payloads, credentials…
Skill by MythicAgents
Chalk Labs
Need help deploying sage-artifact-retention?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/mythicagents-sage-artifact-retention
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md and code. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
- Runs as
- Python package
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/sage-artifact-retention/tmp/claude-/tmp/pip-/tmp/pytest-/tmp/sage_arch_gate/tmp/sage_arch_review/tmp/sage_payloads/tmp/sage-frozen-review-/tmp/sage-isc49-primary-bootstrap./tmp/sage-isc49-review-lease-probe-/tmp/sessions/tmp/tmp
What's inside
- Skill name
sage-artifact-retention- Description
- Preserve high-value Sage contracts, handoffs, reviews, transcripts, and evidence in a private project-local history while keeping payloads, credentials, locks, fixtures, clones, and other scratch work temporary.
- License
- GPL-3.0
- Instructions
- 79 lines in SKILL.md · view on GitHub
Before installing
- Ships executable scripts
Files (6)
- SKILL.md
- openai.yaml · agents/
- artifact_retention.py · scripts/
- retention_guard.py · scripts/
- test_artifact_retention.py · tests/
- test_retention_guard.py · tests/
SKILL.md (first lines)
# Sage Artifact Retention Use this skill whenever a Sage task creates evidence or decision material that must survive a reboot. The durable root defaults to `.sage_history/` at the repository root and can be overridden for testing with `SAGE_HISTORY_ROOT`. It is private, gitignored local state; it is not a system backup. ## Retention Classes - `scratch`: active locks and leases, temporary clones, fixtures, staging, payload downloads, environment snapshots, and reproducible intermediate files. Keep these under `/tmp`. - `durable-private`: contracts, final handoffs, decision-bearing reviews, external panel packets and responses, full chat transcripts used for analysis, accepted or rejected evaluation evidence, closed governance receipts, and manifests. Store these under `.sage_history/`. - `published`: a later curated and redacted subset. Never publish directly from the raw private archive. An artifact is durable when any of these statements is true: - a final response cites it; - it approves, rejects, or materially explains a decision; - it proves an evaluation or live-run result; - it contains a worker contract or final handoff; - resuming accurately after reboot requires its exact bytes. `/tmp` must not be the source of truth for a durable artifact. ## Commands Initialize or inspect the private history: ```bash .venv/bin/python skills/sage-artifact-retention/scripts/artifact_retention.py init .venv/bin/python skills/sage-artifact-retention/scripts/artifact_retention.py manifest ``` Allocate a durable path for a tool that cannot import the helper: ```bash .venv/bin/python skills/sage-artifact-retention/scripts/artifact_retention.py path --category transcripts/native-chat --name request.json ``` After that tool writes the file or directory, append its hashes to the manifest: ```bash .venv/bin/python skills/sage-artifact-retention/scripts/artifact_retention.py record --category transcripts/native-chat --artifact-type native-chat-transcript /path/returned/above ``` Write a structured handoff from a JSON file or standard input: ```bash .venv/bin/python skills/sage-artifact-retention/scripts/artifact_retention.py write-json --category handoffs --name handoff.json --artifact-type task-handoff --input /path/to/handoff.json ``` Copy explicitly reviewed temp artifacts into durable history: ```bash .venv/bin/python skills/sage-artifact-retention/scripts/artifact_retention.py promote --category migrated/reviews --artifact-type external-review --context 'approved retention migration' /tmp/example-review.md ``` Promotion is copy-only. It refuses symlinks, files larger than 100 MiB, and names shaped like credentials, environment files, private keys, or payloads. Do not bypass that refusal by renaming sensitive material. ## Manifest `.sage_history/manifest.jsonl` is an append-only, machine-readable index. Each record binds an artifact or promoted directory to its relative path, SHA-256, size, type, source path when applicable, and recording time. Raw history remains operator-sensitive even when obvious auth files are excluded. ## Completion Guard The project `Stop` and `SubagentStop` hooks run `retention_guard.py`. It scans only a bounded tail of the current transcript, warns about existing high-value `/tmp` paths that lack a matching manifest record, and never copies anything automatically. Treat it as a guardrail, not a complete parser or a substitute for the retention rule. ## Scripts - `scripts/artifact_retention.py`: path allocation, private JSON writes, explicit promotion, and manifest queries. - `scripts/retention_guard.py`: bounded completion-time warning for unpromoted high-value temp artifacts.
From MythicAgents/sage, GPL-3.0.
More from MythicAgents/sage
- 0.655score
sage-trace-analysis
MythicAgents
Repo-local Sage trace, Phoenix, Mythic task-output, and run-log analysis workflow.
SkillAFreshness: A★ 29Claude Code +9 - 0.655score
sage-canary-attestation
MythicAgents
Independently attest a bounded native Mythic canary by reconstructing the run from Mythic's own records with a read-only Spectator credential, then diffing…
SkillAFreshness: A★ 29Claude Code +9 - 0.655score
sage-eval-gauge
MythicAgents
Repo-local Sage eval-gauge / hill-climbing Phase-0 toolkit.
SkillAFreshness: A★ 29Claude Code +9 - 0.655score
sage-focused-capability-tests
MythicAgents
Repo-local focused Sage capability, adapter, and notification validation workflow.
SkillAFreshness: A★ 29Claude Code +9 - 0.655score
sage-foothold-launch
MythicAgents
Build a live Apollo foothold on the GOAD range as NORTH\samwell.tarly through a pure WinRM scheduled-task batch logon.
SkillAFreshness: A★ 29Claude Code +9 - 0.652score
sage-architecture-governor
MythicAgents
Use for Sage prompt, topology, tool, planner, runtime authority, and other consequential architecture changes; keep ordinary development measurement plumbing…
SkillAFreshness: A★ 29Claude Code +9 - 0.652score
sage-callback-bootstrap
MythicAgents
Repo-local post-Mythic-reset Sage chat and foothold bootstrap workflow.
SkillAFreshness: A★ 29Claude Code +9 - 0.652score
sage-live-runner
MythicAgents
Repo-local Sage native Mythic v4 chat execution and inspection workflow.
SkillAFreshness: A★ 29Claude Code +9 - 0.651score
sage-goad-reset
MythicAgents
Repo-local Sage GOAD/Trust Walker full lab reset and readiness workflow.
SkillAFreshness: A★ 29Claude Code +9 - 0.651score
sage-mythic-payload-deploy
MythicAgents
Download an existing Mythic payload, stage it onto a Ludus Windows host over WinRM, and launch it as the current user or a supplied run-as user.
SkillAFreshness: A★ 29Claude Code +9 - 0.650score
sage-forge-ops
MythicAgents
Repo-local Sage Forge/Codex helper workflow.
SkillAFreshness: A★ 29Claude Code +9 - 0.650score
phoenix-traces
MythicAgents
Analyze Sage's Phoenix/OpenInference trace database (phoenix.db) for the Sage LangGraph multi-agent system.
SkillAFreshness: A★ 29Claude Code +9
Signals
- Stars
- 29
- Forks
- 5
- Open issues
- 1
- Last commit (this folder)
- Jul 29, 2026 (3 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- GPL-3.0
- Repository
- github.com/MythicAgents/sage
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from MythicAgents
- 0.655score
sage-trace-analysis
MythicAgents
Repo-local Sage trace, Phoenix, Mythic task-output, and run-log analysis workflow.
SkillAFreshness: A★ 29Claude Code +9 - 0.655score
sage-canary-attestation
MythicAgents
Independently attest a bounded native Mythic canary by reconstructing the run from Mythic's own records with a read-only Spectator credential, then diffing…
SkillAFreshness: A★ 29Claude Code +9 - 0.655score
sage-eval-gauge
MythicAgents
Repo-local Sage eval-gauge / hill-climbing Phase-0 toolkit.
SkillAFreshness: A★ 29Claude Code +9 - 0.655score
sage-focused-capability-tests
MythicAgents
Repo-local focused Sage capability, adapter, and notification validation workflow.
SkillAFreshness: A★ 29Claude Code +9 - 0.655score
sage-foothold-launch
MythicAgents
Build a live Apollo foothold on the GOAD range as NORTH\samwell.tarly through a pure WinRM scheduled-task batch logon.
SkillAFreshness: A★ 29Claude Code +9 - 0.652score
sage-callback-bootstrap
MythicAgents
Repo-local post-Mythic-reset Sage chat and foothold bootstrap workflow.
SkillAFreshness: A★ 29Claude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get sage-artifact-retention running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
sage-artifact-retention skill. Agent Mart. https://agentmart.dev/skills/mythicagents-sage-artifact-retention (last updated 2026-10-24)